GeosocialFootprint (2103): social media location privacy Web map

PDF

Download

Open document

Flip pages

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content i

GEOSOCIALFOOTPRINT(2013):

SOCIAL
MEDIA
LOCATION
PRIVACY
WEB
MAP

by

Christopher
Donald
Weidemann

A
Thesis
Presented
to
the

FACULTY
OF
THE
USC
GRADUATE
SCHOOL

UNIVERSITY
OF
SOUTHERN
CALIFORNIA

In
Partial
Fulfillment
of
the

Requirements
for
the
Degree

MASTER
OF
SCIENCE

(GEOGRAPHIC
INFORMATION
SCIENCE
AND
TECHNOLOGY)

March
2014

Copyright
2013

Christopher
D.
Weidemann

ii

ACKNOWLEDGEMENTS

First
and
foremost
I
offer
my
sincerest
gratitude
to
my
committee
Chair,
Dr.
Jennifer

Swift.
I
will
always
be
grateful
for
the
continued
encouragement
and
support
she

offered
during
this
overwhelming
process.
I
would
like
to
thank
my
committee
as
a

whole,
Dr.
Jennifer
Swift,
Dr.
Edward
Pultar,
and
Dr.
Craig
Knoblock
for
their
time,

resources,
insight,
and
guidance.

I
would
also
like
to
thank
my
family
for
their
constant
support
and
inspiration

throughout
this
process.
I
thank
my
parents
for
always
giving
me
freedom
in
my

endeavors
yet
consistently
pushing
for
success.
Not
only
does
my
wife
deserve
the

credit
for
introducing
me
to
Geographic
Information
Systems
(GIS)
but
now
she
also

deserves
the
credit
in
helping
me
complete
my
thesis
in
GIS.
Without
her,
not
only

would
this
thesis
be
much
worse
off,
but
I
as
an
individual
would
not
have
the

confidence
in
myself
or
the
dedication
to
complete
this
work.

iii

TABLE
OF
CONTENTS

ACKNOWLEDGEMENTS
.........................................................................................................
ii

LIST
OF
FIGURES
.......................................................................................................................
v

List
of
Abbreviations
.............................................................................................................
vi

ABSTRACT
................................................................................................................................
vii

CHAPTER
1:
TOPIC
DEFINITION
AND
BACKGROUND
..................................................
1

1.1
Review
of
relevant
research
....................................................................................
3

1.1.1
Relevant
literature
.....................................................................................................................
3

1.1.2
Relevant
web
applications
......................................................................................................
7

1.2
Motivation
......................................................................................................................................
10

1.3
General
Objective
........................................................................................................................
14

1.4
Software
and
Platforms
...........................................................................................................
14

CHAPTHER
2:
TECHNOLOGY
AND
APPLICATION
DEVELOPMENT
........................
16

2.1
Technology
...................................................................................................................
16

2.1.1
Software
Development
Language
....................................................................................
17

2.1.2
Platform
As
A
Service
(PAAS)
.............................................................................................
19

2.1.3
Application
Programming
Interface
(API)
..................................................................
20

2.2
Application
Development
.......................................................................................
21

CHAPTHER
3:
METHODS
.....................................................................................................
23

3.1
Requirements
Analysis
............................................................................................
23

3.2
Application
Design
....................................................................................................
24

3.2
Software
Development
............................................................................................
25

3.3
Application
Evaluation
Method
............................................................................
25

3.3.1
Bug
Testing
................................................................................................................................
26

3.3.2
User
Survey
.................................................................................................................................
27

3.4Key
Programming
Challenge
..................................................................................
28

3.3.1
Twitter
API
Changes
..............................................................................................................
28

3.3.2
User
Analytics
............................................................................................................................
31

CHAPTHER
4:
RESULTS
.......................................................................................................
32

4.1
Application
Function
................................................................................................
32

4.1.1
User
risk,
alters,
and
suggestions
.....................................................................................
36

4.1.2
Anonymous
Survey
..................................................................................................................
39

4.1.3
Data
Download
.........................................................................................................................
41

4.2
Summary
of
Survey
Results
....................................................................................
43

CHAPTER
5:
Conclusion
......................................................................................................
50

5.1
Main
Advances
............................................................................................................
50

5.2
Next
Steps
.....................................................................................................................
51

5.3
The
Future
of
GIS
and
social
media
.....................................................................
54

iv

References
...............................................................................................................................
57

Appendix
A:
Survey
Results
...............................................................................................
62

v

LIST
OF
FIGURES

Figure
1
-‐
Application
Flow
diagram
..............................................................................................
22

Figure
2
-‐
GeosocialFootprint
(2013)
Landing
Page
...............................................................
32

Figure
3
-‐
View
of
intermediate
web
page
after
a
valid
Twitter
username
has
been

entered
........................................................................................................................................................
33

Figure
4
-‐
Popup
alerting
user
of
an
invalid
Twitter
username
.........................................
34

Figure
5
-‐
Map
results
of
a
high
risk
Twitter
user
....................................................................
35

Figure
6
-‐
Map
results
when
no
tweets
are
geo-‐located
........................................................
35

Figure
7
-‐
Results
of
a
high-‐risk
user,
including
context
alerts
and
suggestions
to

reduce
risk
.................................................................................................................................................
36

Figure
8
-‐
Feedback
request
popup
................................................................................................
39

Figure
9
-‐
Feedback
page
.....................................................................................................................
40

Figure
10
-‐
Data
download
feature
button
..................................................................................
41

Figure
11
-‐
Contents
of
a
CSV
download
opened
in
Microsoft
Excel
................................
42

Figure
12
-‐
Summarized
responses
to
question
one
on
the
user
survey
.......................
45

Figure
13
-‐
Summarized
responses
to
question
two
on
the
user
survey
.......................
46

Figure
14
-‐
Summarized
responses
to
question
three
on
the
user
survey
....................
47

Figure
15
-‐
Summarized
responses
to
questions
four
and
five
on
the
user
survey
...
48

vi

List
of
Abbreviations

Abbreviation
Meaning

API
Application
Programming
Interface

AWS
Amazon
Web
Service

CDN
Content
Delivery
Network

CSS
Cascading
Style
Sheets

CSV
Comma-‐separated
values

DOM
Document
Object
Management

FEMA
Federal
Emergency
Management
Agency

HTML
Hypertext
Markup
Language

HTTP
Hypertext
Transfer
Protocol

JSON
JavaScript
Object
Notation

OOD
Object-‐oriented
Design

PAAS
Platform
as
a
service

REST
Representational
state
transfer

S3
Simple
Storage
Service

SDLC
Software
Development
Lifecycle

UML
Uniformed
Model
Language

UX
User
experience

VGI
Volunteer
Geographic
Information

vii

ABSTRACT

Spatial
thinking
is
an
abstract
term
and
process
in
regards
to
what
most
of
the

general
population
understand.
Many
people
are
not
well
versed
in
geospatial

terminology,
options
of
use,
and
the
location
intelligence
they
unconsciously

disclose
when
using
social
media
outlets.
This
thesis
integrates
a
unique
technical

web
application
with
GIScience
intended
to
illuminate
the
subsequent
effect

location-‐based
data
can
have
on
one’s
personal
privacy,
security,
and
web-‐presence.

An
innovative
new
web
mapping
application
was
built
for
general
public

consumption
that
aggregates
location
data
from
Twitter,
harvests
ambient
location

information,
analyzes
the
captured
data
to
provide
personal
location
intelligence,

and
visualizes
possible
areas
of
interest.
In
addition,
the
research
examines
the

results
of
an
online
voluntary
survey
collected
from
the
users
of
the
application.

Finally,
this
thesis
discusses
how
these
same
techniques
can
be
applied
to
other

social
media
outlets
along
with
potential
opportunities
to
educate
and
inform
the

general
public
more
about
their
social
media
location
privacy.

1

CHAPTER
1:
TOPIC
DEFINITION
AND
BACKGROUND

Social
media
entities
like
Facebook,
Google+,
and
MySpace
store
large
data
sets
of

personal
information
about
their
users.
Not
only
do
they
store
the
information

provided
to
them
through
registration
and
status
updates,
but
users
also
grant
them

access
to
personal
data
through
other
means,
such
as
internet
browser
cookies,

search
history,
and
even
e-‐mail
conversations.
Social
media
entities
store
this
data

with
trust
from
the
user
that
it
will
only
be
used
to
customize
a
user’s
experience.

Social
media
users
should
be
concerned
with
much
more
than
the
data
they
give
to

their
social
media
providers.
Lini
(2012)
found
that
35
percent
of
hiring
managers,

across
a
wide
array
of
disciplines,
have
rejected
an
applicant
based
purely
on

information
they
found
online.
They
also
found
that
Human
Resource
departments

now
screening
new
hires
through
social
media
searches
has
risen
38.4
percent
since

2008.

The
screening
doesn’t
end
with
employers.
Thieves
use
social
media
as
a
tool
for

gathering
intelligence
and
for
picking
their
victims.
Police
in
Brazil
are
very
familiar

with
kidnapping
and
ransoms,
but
in
2010
they
arrested
a
gang
that
used
social

media
to
stalk,
kidnap,
and
maximize
their
ransom
(Fox
News,
2010).
The
gang,
who

has
been
attributed
with
kidnapping
up
to
19
individuals,
first
finds
their
victims

online
through
extravagant
social
media
posts.
They
then
stalk
the
user
physically

2

by
visiting
the
locations
the
user
geo
tags
or
references
in
text.
Finally
the
gang

kidnaps
the
user
and
evaluates
their
worth
through
their
tweets
and
their
social

connections.

Theft
through
social
media
isn’t
restricted
to
Brazil.
In
2008
United
States
vice

presidential
candidate
Sarah
Palin
had
her
e-‐mail
account
stolen
after
a
thief
was

able
to
gather
intelligence
from
her
social
media
outlets
and
web
searches.
The

National
Foundation
for
Credit
Counseling
has
found
that
these
types
of
social

media
tactics
and
online
identity
thefts
are
ground
zero
for
credit
card
identity

thieves
(Benda
2010).
Loeffler
(2012)
argues
that
privacy
laws
need
to
be
updated

to
reflect
evolving
social
media
trends.

These
thefts
are
made
possible
through
the
seemingly
innocent
sharing
of

information
on
social
media
websites.
For
example,
most
Internet
users
don’t
stop

to
think
that
a
mother’s
maiden
name,
which
is
a
common
secondary
security

question
online,
can
be
gathered
from
social
media
relationships.
It
is
also
common

for
users
to
share,
seemingly
harmless
photographs
of
their
families
or
homes.
More

interesting
to
this
thesis,
it
is
very
common
for
social
media
users
to
disclose
their

real-‐time
location
to
what
they
assume
are
their
family
and
friends.
To
the
ill

intentioned,
this
is
all
yet
more
information
that
could
be
used
against
the
social

media
user.

3

1.1
Review
of
relevant
research

1.1.1
Relevant
literature

Between
2005
and
2012
extensive
research
was
been
done
on
the
need
for
privacy

constraints
within
social
networking
(Stefanidis
et
al.
2011,
Friedland
and
Sommer,

2010).
However,
to
date
very
limited
research
has

been
accomplished
related
to
the

subcategory
of
location
privacy
within
social
networks.
Recent
research
is
currently

being
done
on
spatial
data
collection
from
social
media
and
alterative
research

about
location
based
cyber
stalking,
but
no
recent
article
was
found
that
specifically

focused
on
social
media
location
privacy
concerns
or
methods
for
systematically

mining
potential
location
privacy
threats.

In
regards
to
research
similar
to
this
thesis
work,
the
George
Mason
University

Geosocial
research
team
has
proposed
a
system
architecture
for
capturing

geospatial
information
from
social
media
as
volunteer
geographic
information

(VGI),
and
presents
methods
for
analyzing
social
media
streams
for
event
“Hot

Spots”
(Goodchild
2007;
Stefanidis
et
al.
2011).
The
results
of
that
study
validate
the

use
of
social
media
analysis
as
a
data
mining
opportunity
for
different
types
of

alerts,
as
well
as
a
means
to
monitor
real
world
events.
In
addition,
this
study

revealed
the
potential
for
the
harmful
use
of
social
media
data
and
suggested
that

social
media
streams
be
considered
“pseudo”
volunteer
geographic
information

even
through
it
is
defined
as
“ambient”
geospatial
information.
This
ambient
data
is

4

defined
as
any
data
that
references
a
location
yet
isn’t
spatially
located.
In
the
terms

of
relevance
to
this
thesis
work,
the
ambient
geospatial
data
described
in

Humphreys,
L.,
et
al.
(2010)
builds
on
the
idea
of
ambient
data,
but
is
not
described

in
terms
of
spatial
location.
Humphreys
et
al.
(2010)
instead
introduces
the
idea
of

“coded
phrases,”
or
in
other
terms,
using
phrases
in
the
text
of
the
tweet
to
classify

tweets.
In
the
research
of
Humphreys
et
al.
(2010)
coding
tweets
was
classified
in
a

broad
sense
as
activities,
locations,
proper
names,
times
of
day,
and
information

about
the
author
themselves.
This
thesis
and
web
application,
GeosocialFootprint

(2013)
(geosocialfootprint.com,
2013),
merges
this
train
of
thought
with
the

Stefanidis
(et
al.
2011)
definition
of
ambient
location
data
to
classify
locations
as

homes,
places
of
employment,
or
places
of
education.

In
another
study
attempting
to
raise
awareness
about
cybercasing,
Friedland
and

Sommer
(2010)
scrutinize
the
activity
of
using
publically
available
geo-‐information

in
conjunction
with
geo-‐tagged
pictures
and
videos
to
infer
real
world
situational

awareness
for
questionable
purposes.
The
authors
focus
on
geo-‐tagged
pictures
and

video
for
the
source
of
geospatial
information
instead
of
social
media
outlets.
They

argue
that
most
submitters
of
the
geo-‐tagged
media
are
unaware
of
the
location

information
they’re
publishing
as
metadata
and
suggest
steps
need
to
be
taken
on

mobile
devices
to
decrease
privacy
concerns
and
increase
awareness.

5

Additionally,
there
has
also
been
significant
research
accomplished
concerning

opportunities
to
increase
location
privacy
through
computational
awareness.
For

example
Krumm
(2009)
discusses
opportunities
for
situational
based
restrictions

on
location
data.
The
author
argues
that
while
most
people
do
not
seem
to

comprehend
the
potential
negative
consequences
of
sharing
location
data,
system

designers
should
be
responsible
and
instinctively
protect
users’
privacy
through

location
anonymizing
algorithms.

Another
article
by
Barkhuus
L.
(2004)
attempts
to
research,
through
human

subjects,
a
user’s
level
of
concern
with
sharing
location
based
data.

In
their
human

subject
research
they
performed
two
tests.
In
the
first
test
they
had
subjects
carry
a

pseudo
location-‐tracking
unit
and
then
record
their
location
concerns
at
the
end
of

each
day
in
a
journal.
The
second
experiment
was
designed
as
a
real
world
example

of
sharing.
It
involved
supplying
another
set
of
students
with
location
tracking

devices
and
asking
them
to
share
their
location
as
they
saw
fit.
The
students
were

not
required
to
journal
about
their
experience.
Therefore
they
were
not
required
to

reflect
on
their
concerns.
Barkhuus
found
that
while
their
research
subjects
initially

showed
concern
for
sharing
their
location
in
the
first
experiment,
the
concern
was

drastically
decreased
in
the
second
experiment
when
reflection
was
not
forced
upon

the
subject.
Their
study
is
limited
due
to
the
fact
that
nearly
all
their
subjects
fall

within
one
demographic,
college
students,
but,
nonetheless,
it
proves
that
at
least

one
subset
of
people
do
not
seem
to
care
about
their
location
privacy.

6

This
thesis
argues
that
while
the
relationships
and
photographs
shared
through

social
media
should
be
of
concern
to
the
user,
they
should
be
equally
aware
of

location
privacy
and
the
potential
risk
they
place
themselves
in
due
to
their
location

sharing.
This
thesis
advances
this
concern
by
building
upon
recent
research
and

arguments
made
by
several
authors
discussed
in
the
background
review
on
this

topic,
including
Stefanidis
et
al.(
2011),
Friedland
and
Sommer,
2010,
Barkhuus
L.

(2004),
Krumm
(2009),
and
Humphreys,
L.,
et
al.
(2010).

The
application
developed
and
reported
in
this
thesis
is
unique
as
it
attempts
to

shed
light
on
the
ease
with
which
personal
location
data
can
be
extracted
and

analyzed
for
privacy
concerns.
It
provides
the
user
with
a
dynamic
experience
that

not
only
informs
them
of
such
privacy
concerns,
but
also
provides
feedback
for

decreasing
such
risks.

While
the
application
focuses
on
Twitter,
these
same

techniques
can
be
adapted
to
collect
and
analyze
other
social
media
data.
In
turn

such
value-‐added
information
can
raise
awareness
to
both
social
media
users
and

providers
regarding
weaknesses
as
well
as
strengths
in
the
privacy
options
offered

to
consumers.

7

1.1.2
Relevant
web
applications

In
recent
years
there
have
only
been
a
few
web
applications
that
approximated
the

functionality
of
the
web
application,
GeosocialFootprint
(2013),
developed
as
part
of

this
thesis
work.
WeKnowYourHouse.com
(2011)
and
PleaseRobMe.com
(2012)

both
highlight
known
possible
unintended
side
effects
of
sharing
your
location
on

social
media.
There
are
also
other
tools,
like
Tweography
(2012),
that
let
users
map

their
own
twitter
stream.
Finally,
there
are
also
web
applications
that
allow
users
to

map
and
monitor
the
entire
Twitter
Stream
as
a
whole,
such
as
Tweetping.net

(2012).

WeKnowYourHouse.com
(2011),
which
was
taken
offline
in
2011
due
to
legal
action

from
Twitter,
actively
scanned
Twitter
for
geo-‐enabled
tweets
that
made
a
textual

reference
to
a
user’s
home.
From
there,
the
web
application
mapped
the
location

using
Google
Streetview
and
automatically
responded
to
the
user
on
Twitter.
While

all
the
information
the
web
application
used
was
public,
in
late
2011
the
application

owners
received
a
takedown
request
from
Twitter.com
due
to
privacy
concerns

over
the
automated
tweets
providing
estimated
address
locations
for
selected
users.

WeKnowYourHouse
received
a
wide
array
of
media
coverage
due
to
its
extreme

stance
and
proactive
nature
in
exposing
location
privacy
concerns
on
Twitter.

GeosocialFootprint
(2013)
employees
similar
ambient
location
context
matching,

but
then
builds
upon
the
existing
work
by
providing
additional
analysis
and

visualization.

8

A
second
application
that
has
received
significant
media
coverage
due
to
social

media
over-‐sharing
is
PleaseRobMe.com
(2012).
Instead
of
searching
for
people’s

homes,
it
does
the
near
opposite
and
allows
a
user
to
enter
their
twitter
screen

name
and
see
their
foursquare
check-‐in
activity,
and
therefore
shows
when
a
user
is

not
at
home.
It
expands
the
definition
of
personal
privacy
web
tools
and
has

received
significant
media
coverage
because
of
it.

While
the
application
is
simple

and
provides
no
location
context,
it
serves
as
a
reminder
to
all
users
that
the
twitter

data,
and
any
location
data
associated
with
it,
is
public.
Regardless
of
what
the
name

implies,
the
site
actively
helped
Twitter
users
from
disclosing
personal
information

by
alerting
potential
targets
of
their
over-‐sharing
habits.
Unfortunately
due
to

changes
in
Twitter’s
application
programming
interface
(API),
PleaseRobMe
(2012)

no
longer
works.

While
Tweography.com
(2011)
has
also
failed
to
evolve
with
the
recent
Twitter
API

changes,
it
still
provides
a
solid
example
of
the
value
of
geographic
visualization.

Tweography
simply
displayed
all
of
a
user’s
geo-‐enabled
tweets
on
a
map.
Unlike

GeosocialFootprint
(2013),
Tweography
does
not
provide
analysis,
risk
accounting,

text
geocoding,
or
educational
features.
However,
they
received
constant
media

coverage
due
to
the
simple
approach
of
helping
users
visualize
their
tweets
on
a

map.
GeosocialFootprint
(2013)
does
build
upon
the
concept
of
simple
tweet

visualization
and
reaped
the
benefits
of
extensive
media
coverage
as
well.

9

While
each
of
the
sites
listed
previously
in
this
section
were
unique
in
their

application,
there
is
another
series
of
twitter
applications
that
focus
on
the
macro

visualization
of
tweets.
Instead
of
focusing
on
one
user’s
tweets,
Tweetping.net

(2012)
visualizes
all
live
tweets
on
a
map.
What
it
lacks
in
micro
and
privacy
focus,

it
makes
up
for
in
analysis.
Tweetping
expands
on
the
simple
geographic

visualization
of
Tweography
and
adds
an
analysis
overlay,
which
provides
some

simple
metrics
for
and
key
indicators.
These
allow
the
user
to
correlate
the
visual

representation
on
the
map
with
the
actual
data
found
in
the
tweets.
Tweetping

provides
an
example
that
validates
ways
to
provide
meta
data
and
simple
statistics

on
the
web
application
to
greatly
increase
its
functionality.

Weknowwhatyouredoing.com
(2011)
is
an
educational
web
site
listed
as
a
social

networking
privacy
experiment.
The
application
focuses
on
the
entire
Twitter

stream,
not
just
individual
users.
It
sorts
and
filters
tweets
based
of
key
phrases
that

are
expected
to
divulge
some
type
of
personal
or
identifiable
information.
Their

main
page
provides
a
stream
of
current
Twitter
users
who,
based
off
the
context
of

their
tweets,
are
either
hung
over,
taking
drugs,
speaking
poorly
about
their
boss,
or

have
recently
published
a
new
phone
number
on
Twitter.
They
also
provide

secondary
searches
for
users
that
have
used
Four
Square,
a
check
in
service,
to

inadvertently
provide
coordinates
for
their
“home.”
This
web
site
tries
to
shed
light

10

on
the
privacy
risks
assumed
through
over-‐sharing
but
does
not
provide
any
type
of

location
analysis
or
advanced
web
mapping
tools
such
as
heat
mapping.

Since
the
release
of
GeosocialFootprint,
UC
Berkley’s
Teaching
Privacy
group
has

released
an
application
that
allows
users
to
visualize
their
twitter
footprint
named

Ready
or
Not.
It
also
performs
a
temporal
analysis
of
the
user’s
tweets
which
can

show
patterns
in
weekly
tweeting
habits.
The
group’s
application
provides
no

additional
analysis
or
visualization.

While
each
of
these
relevant
web
applications
serves
a
solid
purpose
in
their
own

right,
none
of
them
explicitly
allow
a
Twitter
user
to
visualize
their
geo-‐enabled

activity,
view
possible
risks,
highlight
areas
of
concern,
and
provide
dynamic

instructions
on
how
to
reduce
their
over-‐sharing
habits.
This
thesis
and
application

merge
many
aspects
and
ideas
found
in
these
previously
referenced
web

applications
into
a
new
tool
that
fills
the
gaps
left
behind
in
serving
this
projects’

desired
purpose.

1.2
Motivation

The
research
completed
by
Stefanidis
(et
al.
2011)
on
ambient
location
information

was
inspiring
in
that
it
provided
a
methodological
foundation
for
developing
a

systematic
approach
to
ambient
location
mining.
While
the
data
in
that
study
was

11

used
to
extract
macro
level
information
on
world
events,
the
same
theory
can
be

applied
to
the
individual
user
and
therefore
their
personal
privacy
concerns.

Other

researchers
have
inadvertently
used
some
of
this
information
to
define
location,
but

Stefanidis
(et
al.
2011)
used
it
as
an
exclusive
data
source.
This
thesis
and

GeosoicalFootprint
(2013)
have
been
adopted
from
the
same
train
of
thought;

ambient
data
should
be
considered,
at
the
very
least,
valid
metadata
worthy
of

analysis.

While
Friedland
and
Sommer
(2010)
may
argue
that
software
developers
should

instinctively
anonymize
the
geo-‐tagged
data,
this
thesis
work
takes
an
entirely

unique
and
courteous
approach
of
intending
to
inform
only
the
individual
end
user

instead.
With
that
being
said,
this
previous
work
illuminating
the
twitter
general

user
population’s
lack
of
awareness
of
the
dangers
of
geosocial
over-‐sharing
truly

validated
the
need
for
this
thesis
work
and
thus
development
of
this
web

application.

Barkhuus’
(2004)
work
again
validates
the
need
for
an
application
that
educates
the

end
user
on
the
risks
of
location
over
sharing.
This
thesis
and
web
application

serves
to
fill
the
need
of
location
privacy
education
through
data
visualization
for

Twitter
users.
Additionally
it
will
provide
opportunities
for
users
to
reflect
on

location
privacy
risks
by
offering
basic
alerts,
highlighting
areas
of
concern,
and

customized
educational
opportunities
through
third
party
resources
on
over-‐

12

sharing.
Their
human
subject
research,
along
with
the
conclusion
they
have
drawn

was
a
solid
foundation
for
building
my
own
human
subject
research.

A
strong
motivation
for
this
research
is
the
fact
that
no
web
application
currently

exists
that
allows
a
user
to
view
and
manage
their
social
media
location
privacy

concerns.
The
other
projects
mentioned
above
outline
specific
privacy
concerns

divulged
through
social
media,
but
are
not
full
monitoring
solutions.

One
of
the
first
applications
that
sparked
my
interest
for
this
topic
was

WeKnowYourHouse.com
(2011).
The
strong
media
coverage
it
received
seemed
to

be
thought
provoking
for
its
readers.
In
fact,
it
was
so
impactful
that
Twitter
had
to

step
in
and
request
that
the
service
be
taken
down.
It
serves
as
a
good
example
for

this
thesis
in
that
it
takes
an
extreme
stance
and
innovation
to
gain
media
attention,

yet
a
strict
adherence
to
all
rules
and
policies
to
maintain
an
extended
impact.

PleaseRobMe
(2012)
shows
again
that
a
unique
and
bold
application
is
required
to

catch
the
attention
of
users
and
the
media.
Through
the
media
attention
a
large

impact
can
be
made
on
over
sharing.
In
addition
it
can
be
noted
from
the

experiences
of
PleaseRobMe
(2012)
that
API
evolution
must
be
considered
and

accounted
for
when
designing
a
web
application.
The
web
application
in
this
thesis

was
designed
with
such
API
changes
in
mind
and
will
easily
accommodate
future

changes
in
Twitter’s
API
endpoints.

13

The
simple
design
of
Tweography
(2011)
provided
a
great
example
to
base
the

design
of
this
thesis
and
project
after.
With
the
understanding
that
many
users
were

educated
through
the
simple
visualization
of
tweets,
during
the
development
of

GeosocialFootprint
(2013),
it
was
important
that
Tweography’s(2011)
idea

mirrored
and
then
built
upon.

The
idea
of
classifying
a
tweet
based
off
the
text,
which
is
also
referenced
in

Humphreys,
L.,
et
al.
(2010),
is
a
strong
feature
of
GeosocialFootprint
(2013)
(2013)

and
was
built
off
the
ideas
found
in
both
Weknowwhatyouredoing
(2012).

While

Weknowwhatyouredoing
(2012)
focuses
on
the
macro
level
of
Twitter,
the
same

principles
of
classifying
text
based
on
phrases
can
be
applied
to
the
individual
user.

Instead
of
being
interested
in
what
users
are
using
drugs,
this
thesis
focuses
on

classifying
users
locations.

Tweetping
(2012)
is
an
example
and
validation
for
how
providing
metadata
and

simple
statistics
on
the
web
application
can
greatly
increase
it’s
functionality.
The

application
developed
for
this
thesis
does
not
provide
analysis
on
the
macro
level
or

even
access
to
the
live
tweet
stream,
however
it
does
mirror
some
of
metadata

functionality
in
its
risk
calculation
and
areas
of
concern.

14

1.3
General
Objective

GeosocialFootprint
(2013)
aims
to
help
inform
social
media
users
of
their
personal

social
media
location
footprint
and
the
potential
risks
they
are
susceptible
to

through
social
media
over
sharing.
This
is
done
through
merging
successful
aspects

of
the
pertinent
applications
listed
above
and
implementing
the
research
found
in

the
relevant
literature.

The
application
was
built
with
a
simple
user
interface
and
experience
in
mind.
It

was
assumed
that
advanced
technical
users
would
be
more
aware
of
the
risks
of

over
sharing,
therefore
the
interface
needed
to
accommodate
for
novice
and

intermediate
users.
This
is
accomplished
through
simple,
intuitive
design
and

clearly
defined
instructions.

Based
on
the
experience
of
similar
projects,
specificallyWeKnowYourHouse.com

(2011),
precautions
were
taken
to
minimize
the
risk
of
legal
action
from
entities
like

Twitter.
GeosocialFootprint
(2013)
complies
with
all
known
understandings
of
the

Twitter.com
and
Twitter
API
terms
of
use
and
such
terms
of
use
were
implicitly

followed
during
the
design
and
development
phase
of
the
application.

1.4
Software
and
Platforms

JavaScript
was
selected
as
the
software
development
language
of
choice
for
this

project.
JavaScript
allows
for
cross
platform
use
and
availability,
including
desktop

15

and
mobile
computing(Microsoft
MSDN
2013).
JavaScript
also
allows
for
the
use
of

client
side
application
logic,
which
enables
the
analysis
to
be
performed
on
the

user's
computer
instead
of
a
central
server.
This
in
turn
reduces
hardware
needs.

Due
to
the
use
of
client
side
logic,
the
web
application
is
hosted
on
Amazon’s
S3

cloud
services.
A
proxy
has
also
been
built
on
top
of
Heroku’s
(2013)
service

oriented
infrastructure
that
allows
for
user
interaction
with
Twitter’s
API.
Both
of

these
services
allow
for
elastic
use,
meaning
the
hosting
environments
react
to

client
requests.
Therefore,
hosting
costs
are
reduced
as
site
usage
is
reduced.

Other
software
development
languages
and
hardware
hosting
options
are
available

for
use,
but
these
selections
represent
emerging
computational
trends
barely
used

in
the
previous
studies
cited
above.
Consequently
GeosocialFootprint
(2013)
is
not

only
innovative
in
its
analysis,
but
also
its
technologic
use.

16

CHAPTHER
2:
TECHNOLOGY
AND
APPLICATION
DEVELOPMENT

2.1
Technology

Technology
is
always
evolving.
As
technology
evolves,
so
do
the
uses
of
the
data

collected
and
stored
by
these
new
technologies.
Both
the
geospatial
and
social

networking
industries
have
capitalized
on
the
evolution
of
technology.
The

geospatial
industry
has
continued
to
expand
its
analytical
capabilities
to
reach
new

mediums,
such
as
the
web,
and
improved
its
data
collection
proficiencies.
Social

media,
among
other
things,
has
begun
to
augment
its
data
collection
habits
with

spatial
information.

As
such,
there
are
opportunities
for
these
technologies
to

overlap.
GeosocialFoorprint.com
serves
to
be
a
melding
of
both
industries’

technologies
as
it
extends
the
location
based
information
collected
by
social

networks
to
a
web
based
geospatial
application.

While
the
main
objective
of
GeosocialFootprint
(2013)
is
to
educate
and
inform

social
media
users
of
potential
location
over
sharing,
a
subset
of
technical
design

objects
were
put
in
place
to
focus
and
formalize
the
application
development
efforts.

The
subset
of
objectives
includes:
simple
user
interface,
client
side
application

processing,
open
source
technologies,
and
web
application
event
tracking.

As
the
previous
research
depicts,
a
simple
user
interface
can
provide
an
improved

user
experience
for
the
untrained.
Client
side
logic
reduces
the
overall
hosting
costs

17

and
improves
the
applications
response
time
by
allowing
the
user’s
Internet

browser
to
handle
the
data
storage
and
processing.
Building
upon
and
further

developing
open
source
technologies
not
only
expands
the
body
of
knowledge
but

also
reduces
to
cost
of
entry
for
this
applications
and
others
like
it.
Selective
web

application
event
tracking
will
allow
for
thesis
validation.

This
thesis
project
relies
heavily
on
the
use
of
third
party
API’s
to
access
data
along

with
third
party
libraries
to
analyze
and
display
results.
Google
Maps
API
version

3.3
is
used
as
the
web-‐mapping
base.
Twitter’s
API
is
used
to
access
the
tweets.

Bootstrap.js
and
jQuery.js
libraries
were
both
implemented
to
efficiently
develop

the
web
application.

2.1.1
Software
Development
Language

Since
client
side
application
processing
is
a
fundamental
goal
of
the
application,
the

large
majority
of
Geosocialfootprint.com
was
developed
using
HyperText
Markup

Language(HTML),
Cascading
Style
Sheets
(CSS),
and
JavaScript.
Just
like
this
thesis

builds
upon
much
of
the
research
that
has
come
before
it,
the
web
application
is
also

built
upon
existing
open
source
projects
that
utilize
such
coding
languages.

In
the
setting
of
GeosocialFootprint.com,
JavaScript
is
used
not
only
to
collect,

process,
analyze
and
visualize
the
tweets,
but
it’s
also
used
as
a
mechanism
to

manage
the
dynamically
changing
HTML
code
within
the
application.
This
is
also

18

referred
to
as
Document
Object
Model
(DOM)
manipulation.

Bootstrap.js
(2013)

andjQuery
(2013)
both
provide
for
an
agile
web
application
platform
while
using

JavaScript
and
DOM
manipulation.
They
are
preconfigured
templates
of
HTML,
CSS,

and
JavaScript
code
that
can
be
combined
with
custom
web
applications
to
improve

functionality
and
visualization.

Bootstrap.js
controls
much
of
the
layout
and
design

of
the
application:
headings,
buttons,
input
boxes,
and
etc.
Among
other
things,

jQuery
provides
a
visual
enhancement
for
elements
such
as
feedback
popups,

transitions,
and
etc.

Not
only
do
these
frameworks
allow
for
efficient
development
but
they
also
mimic

accepted
industry
standards
for
user
interface
design
and
user
experience

expectations
(Paddock
and
Peterson
2012).
Additionally,
both
implement
elements

of
responsive
design,
which
allows
the
application
to
detect
the
user’s
computing

environment
and
adjust
to
fit
user
hardware
restraints.
This
allows
the
application

to
be
used
on
all
platforms,
including
mobile.

While
jQuery
is
very
efficient
at
DOM
manipulation
and
provides
many
visually

appealing
enhancements,
its
most
utilized
tool
in
GeosocialFootprint
(2013)
is

JavaScript
Object
Notation
(JSON)
management.
jQuery
has
a
series
of
built
in
JSON

centered
functions
that
allow
for
easy
access
and
retrieval
of
Twitter’s
API
data.
Due

to
the
efficiency
of
jQuery’s
JSON
handling,
not
only
is
the
Twitter
API
data
received

19

as
JSON,
but
the
application
developed
for
this
thesis
is
programmed
to
use
JSON
as

a
local
storage
container
for
all
analytical
data
and
results.

2.1.2
Platform
As
A
Service
(PAAS)

Foundationally,
the
platform
used
to
host
a
web
application
is
just
as
important
as

the
code
that
is
executed
for
the
user.
Without
a
reactive
and
efficient
platform,
even

the
best
of
code
can
seem
slow
and
unresponsive.
With
that
understanding,

GeosocialFootprint
(2013)
employs
Platforms
As
A
Service
(PAAS)
in
an
effort
to

maintain
an
efficient
hosting
environment.
Amazon’s
AWS
services
were
selected

for
this
project
due
to
their
service
record,
flexibility,
and
relatively
low
operating

costs.
In
addition
to
Amazon’s
offered
platforms,
two
additional
services
are

operated:
Heroku
(2013)
and
Qbaka
(2013).

Amazon’s
Simple
Storage
Service
(S3)
acts
as
the
physical
web
host
for
the

application
(Barr,
J.,
Tetlaw,
A.,
&
Simoneau,
L.
2010).
As
a
result,
Amazon’s
AWS

environment
manages
the
domain
and
Domain
Name
System
(DNS)
handling.
Due

to
the
S3’s
safe
storage
redundancy
practices
and
their
Content
Delivery
Network

(CDN),
users
of
GeosocialFootprint
(2013)receive
low
latency
and
fast
response

times.

Amazon's
S3
platform
was
selected
due
to
the
dynamic
scalability
of
their

platform
and
service(Wang
et
al.
2010).
Amazon
S3
can
support
the
hosting
of
static

files
across
Amazon's
content
delivery
network
which
enables
dynamic
scaling

using
Amazon's
existing
infrastructure.
This
came
in
particularly
handy
when
news

20

media
outlets
published
stories
on
GeosocialFootprint
(2013)
and
traffic
increased

to
thousands
of
users
per
day.

Heroku
(2013)
is
a
small
PAAS
that
itself
is
built
on
top
of
Amazon’s
AWS.
Through

extending
and
simplifying
the
AWS
functionality,
Heroku
(2013)
makes
it
possible

to
quickly
deploy
server
resources
through
a
template
system.
Heroku
(2013)

Developer’s
write
templates
that
can
be
shared
on
popular
coding
repositories,
like

GitHub,
and
then
directly
deployed
to
a
Heroku
environment.
GeosocialFootprint

(2013)
utilizes
such
a
template
to
handle
the
authentication
and
interaction

between
Twitter
Developer’s
API
and
the
users
web
browser.

While
Qbaka’s
(2013)
service
was
originally
designed
to
help
developers
of
client

side
applications
monitor
errors
received
by
users,
in
GeosocialFootprint
it
is
being

used
to
monitor
and
manage
user
events.
After
the
application
performs
the
analysis

on
the
tweets,
data
is
sent
to
Qbaka
(2013)
that
contains
the
anonymous
event

information.

That
information
has
been
used
in
this
thesis
to
validate
the

applications
effectiveness.

2.1.3
Application
Programming
Interface
(API)

API’s
specify
a
defined
protocol
for
interacting
with
data
or
processes
stored
by

differing
computer
systems.
In
the
context
of
this
thesis,
API’s
provide
access
to

21

social
media
data
from
Twitter,
mapping
base
layers
from
Google
Maps,
and
event

error
management
from
Qbaka.

Google
Maps
API
provides
a
very
nimble
and
responsive
JavaScript
API.
When
used

in
conjunction
with
bootstrap.js
and
jQuery,
it
provides
a
very
suitable
foundation

and
viewing
platform
for
the
spatial
analysis
results.
ESRI’s
JavaScript
API
(ESRI,

2013)
and
Leaflet
(Leaflet,
2013)
were
also
considered
as
options
for
use,
at
the

time
Google
Maps
API
provided
the
most
support
for
the
Heatmap
functionality
in

the
form
of
both
official
Google
documentation
and
user
feedback
through
blogs
and

forums.
Google’s
Map
API
establishes
the
user
with
a
connection
for
base
layer
map

tiles
and
data
heat
map
visualization.

Twitter’s
API
allows
for
the
access
of
data
from
user
profiles,
tweets,
and
queries.

For
the
context
of
this
thesis,
Twitter’s
API
is
used
to
aggregate
a
user’s
most
recent

200
tweets
for
processing
by
GeosocialFootprint
(2013).
An
alternative
to
Twitter’s

API
is
GNIP’s
API
(Gnip
2013),
however
the
latter
is
a
paid
redistributor
of
social

media
and
would
have
provided
no
true
value
added
to
this
thesis.

2.2
Application
Development

In
total,
10
modules
were
created
to
move
the
data
from
Twitter,
parse,
convert,

store,
geoprocess,
analyze,
and
display
results.
The
code
is
executed
through
a
web

22

application
that
inputs
the
user’s
Twitter
ID.
This
project
flows
through
the
common

stages
of
analysis;
data
collection,
data
preparation,
data
analysis,
and
displaying

results.
This
is
depicted
in
the
flow
chart
provided
in
Figure
1
(
see
below):

Figure
1
-‐
Application
Flow
diagram

23

CHAPTHER
3:
METHODS

GeosocialFootprint
(2013)
was
created
using
an
iterative
Software
Design
Life
Cycle

(SDLC,
University
of
California,
San
Francisco,
2013).
Based
on
the
findings
of
the

technology
research
described
in
Chapter
Two,
requirements
for
developing
this

application
were
documented.
Following
the
requirements
analysis
phase,
a
visual

wireframe
of
the
application
along
with
object-‐oriented
designs(OOD)
documents

were
drafted.
A
visual
wireframe
is
essentially
a
visual
design
guide
that
outlines
the

general
layout,
design,
and
functionality
of
a
web
page.
The
OOD
patterns
were
used

to
begin
application
development.
The
development
process
occurred
in
small

iterations
with
phases
of
testing
completed
at
each
milestone,
the
details
of
which

are
provide
in
the
proceeding
sections.

3.1
Requirements
Analysis

During
the
requirements
analysis
phase
of
the
SDLC
two
areas
of
focus
were

selected
for
the
GeosocialFootprint
(2013)
project:
functional
requirements
and

architectural
requirements.
In
the
first
phase,
functional
requirements
gathering,

necessary
user
experience
(UX)
tasks,
actions,
and
activities
were
documented.
In

the
second
phase,
architectural
requirements
gathering,
key
system
architectures

were
identified.

24

Functional
requirements
gathering
for
GeosocialFootprint
(2013)
focused
on
the

desired
UX.
During
this
sub-‐phase
the
emphasis
was
placed
on
defining
a
simple

user
interface,
outlining
industry
standard
expected
user
experiences,
and
any

additional
functionality
required
outside
of
visualization
of
tweets.

Fully
defining
system
architecture
is
important
with
any
technical
development

project.
With
GeosocialFootprint
(2013),
key
architectures
were
outlined
that

allowed
for
client-‐side
computing
logic
as
originally
intended.

By
identifying
key

system
components
early,
technical
hurdles
were
efficiently
addressed
during
the

second
phase
of
SDLC,
application
design.

Trello(2013),
a
collaborative
Kanban
style
project
management
tool
was
used
to

manage
all
GeosocialFootprint
(2013)
requirements.
It
was
selected
due
to
its

simple
user
interface
at
no
cost.
It
provided
a
hierarchal
view
of
project
tasks

through
all
phases,
from
design
to
final
testing.
Within
Trello
(2013),
tasks
were

grouped
together
to
create
coding
iterations
and
milestones
were
defined
for
each

group.

3.2
Application
Design

GeosocialFootprint
(2013)
was
designed
using
web
application
mockup
templates

in
Adobe
Photoshop,
and
also
OOD
design
templates
in
UMLet
(2013).
Due
to
the

utilization
of
the
popular
Bootstrap
framework
and
JQuery
library,
it
was
most

25

convenient
for
visual
wire
framing
to
be
completed
using
open
source
Adobe

Photoshop
templates(Bent
Design
Studios,
2012).
These
templates
allow
for
rapid

visual
prototyping.
During
the
visual
design
phase
of
GeosocialFootprint
(2013),

focus
was
placed
on
a
simple
user
interface.

OOD
documents
were
created
using
Unified
Modeling
Language
(UML)
in
the
UMLet

application.
UML
diagrams
provided
awell-‐organized
approach
to
the
modeling
the

objects,
methods,
and
attributes
of
GeosocialFootprint
(2013).

3.2
Software
Development

As
referenced
in
Chapter
two,
a
variety
of
software
development
languages
were

utilized
in
Geospatial
Footprint.
Sublime
Text
2,
an
advanced
code
editor,
effectively

manages
cross-‐coding
environments
and
therefore
was
utilized
for

GeosocialFootprint
(2013)
software
development
(Sublime
Text
2,
2012).
While
the

four
main
industry
leading
Internet
browsers,
Microsoft
Internet
Explorer,
Mozilla

Firefox,
Google
Chrome
and
Apple
Safari
were
used
in
testing,
Google
Chrome’s

Developer
Tools
were
used
exclusively
during
the
creation
of
the
application
to

monitor
error
reporting
in
an
Internet
browser
setting.

3.3
Application
Evaluation
Method

GeosocialFootprint
(2013)
was
evaluated
for
both
its
efficacy
in
educating
a
user
on

the
potential
risks
of
over-‐sharing
and
also
for
software
bugs.
Bug
testing
measured

26

the
success
of
the
software
development.
User
feedback
was
collected
through
a

user
survey
and
provided
metrics
to
gauge
the
educational
value
and
effectiveness

of
the
application.

3.3.1
Bug
Testing

Application
bug
testing
for
GeosocialFootprint
(2013)
was
accomplished
in
two

different
phases:
unit
testing
and
system
testing.
Unit
refers
to
a
processes
of
testing

individual
sections
of
code,
while
system
refers
to
entire
application
testing.
Due
to

the
iterative
application
development
cycle,
unit
testing
occurred
at
the
conclusion

of
each
development
milestone.
Once
each
unit
had
been
developed
and
tested

individually
an
entire
system
test
was
completed.

Due
to
the
varying
structures
and
objectives
of
each
development
milestone,
each

unit
test
phase
was
implemented
uniquely.
In
general
sample
data
was
stored
and

manually
run
through
the
code
segments
to
ensure
that
calculations
and
logic
were

operating
as
intended.
Bugs
were
documented
in
the
project
management
tool
and

new
tasks
were
created
and
prioritized.
Once
each
unit
was
tested,
the
bugs
were

addressed
and
resolved.
A
second
phase
of
testing
was
then
completed
on
the
same

unit
and
the
cycle
was
repeated
until
all
bugs
were
corrected.

System
testing
began
after
each
unit
had
been
individually
developed
and
tested.

Two
groups
of
beta
testers
were
used
to
perform
system
testing
in
phases.
Non

27

geospatial
professional
peers
were
selected
to
participate
as
beta
testers.
Group
one

was
provided
with
in
depth
instructions
and
sample
data
for
use.
The
testing
focus

of
group
one
was
data
analysis
accuracy.
After
group
one
completed
their
testing

they
were
asked
to
provide
descriptions
of
any
bugs
they
encountered.

The
bugs
documented
during
phase
one
testing
were
corrected
prior
to
initiating

phase
two
testing.
Phase
two
system
testing
focused
on
the
UX.
Group
two
was

provided
with
minimal
instructions
and
no
sample
data.
Testers
were
provided
with

an
unstructured
feedback
request
in
an
effort
to
capture
all
UX
inefficiencies.
These

bugs
and
inefficiencies
were
corrected
prior
to
final
release
of
the
application.

3.3.2
User
Survey

While
the
completeness
and
accuracy
of
GeosocialFootprint’s
(2013)
application

code
was
important,
the
true
measure
for
success
was
its
ability
to
educate
and

inform
the
users
of
over-‐sharing
risks.
To
evaluate
this
success
a
user
survey
was

offered
to
the
public
online
users
on
a
volunteer
basis.

After
two
minutes
of
use,
the
web
application
triggered
a
pop-‐up
window
that

prompted
the
user
to
participate
in
a
user
survey.
The
user
survey
provided
an

opportunity
for
feedback
in
5
areas
that
measure
educational
value
and
one

additional
field
for
general
comments.
For
the
purposes
of
this
thesis
work,
a
total
of

28

352
surveys
were
collected
over
a
three-‐month
period,
from
September
2013

to

January
2014.

3.4Key
Programming
Challenge

GeosocialFootprint
(2013)
encountered
two
development
challenges:
changes
to
the

Twitter
API,
and
capturing
meaningful
user
analytics.
Both
obstacles
were
overcome

by
using
third
party
platforms
to
accommodate
the
design
requirements
of
client-‐
side
computing
logic,
detailed
in
the
following
discussions.

3.3.1
Twitter
API
Changes

An
early
proof
of
concept
for
GeosocialFootprint
(2013)
was
created
using
the

Twitter
API
version
1.0
and
Google
Maps
API
version
2.

Between
the
proof
of

concept
and
the
beginning
of
application
development,
Twitter
depreciated
version

1.0
of
their
API
and
enforced
the
use
of
version
1.1
and
Google
depreciated
version
2

and
enforced
the
use
of
version
3.

While
the
numerical
change
in
API
versions
may

seem
minimal,
the
technical
ramifications
were
significant.

The
Google
Map's
API
version
2
depreciation
made
it
difficult
to
adopt
the

techniques
of
Weidemann
and
Swift
2013
in
regards
to
geocoding
tweets.
Version
2

of
Google’s
goecoding
API
performed
entity
extract,
natural
language
processing,

29

and
an
accuracy
assessment
for
the
API
user.
When
these
functionalities
were

depreciated,
geocoding
tweets
directly
through
Google's
API
became
impractical.

The
three
largest
modifications
to
the
Twitter
API
were:
required
authentication
on

every
API
transaction,
a
change
to
the
rate-‐limiting
methodology,
and
an
adaption
of

developer
rules
with
regards
to
third
party
Twitter
applications.
Each
modification

impacted
GeosocialFootprint
(2013)
differently.
Currently,
the
change
in
rate-‐
limiting
actually
improves
the
usability
of
GeosocialFootprint
(2013)
by
expanding

the
simultaneous
user
limit
from
350
to
720
(Sippey
2013).
Although
the
adaptation

of
developer
rules
has
limited
impact
on
the
current
use
of
GeosocialFootprint

(2013),
it
does
limit
the
ability
to
further
expand
the
application
to
a
native
mobile

or
desktop
application.
This
is
accomplished
by
requiring
preauthorization
from

Twitter
before
distributing
installable
Twitter
application
clients.

Lastly,
requiring

authentication
to
access
an
API
endpoint
had
the
largest
impact
on

GeosocialFootprint
(2013).

While
using
Twitter’s
API
version
1.0,
the
proof
of
concept
application
accessed
the

Twitter
API
through
a
simple
representational
state
transfer
(REST)
request
using

an
unsecured
HyperText
Transfer
Protocol
(HTTP).
With
version
1.1,
each
REST

request
requires
the
inclusion
of
an
authentication
token
and
key.
These
tokens
and

keys
are
unique
to
each
Twitter
user
and
therefore
must
be
secured
and
protected.

30

In
the
context
of
developing
GeosocialFootprint
(2013),
the
Twitter
API
version
1.1

presented
the
challenge
of
protecting
the
authentication
token
and
key
while

simultaneously
providing
a
user-‐friendly
experience.
Two
options
to
overcome
this

challenge
were
evaluated:
1)
require
each
user
to
login
using
Twitter’s
“OAuth”

protocol
and
2)
utilize
an
API
proxy.

Evaluating
the
first
challenge
of
implementing
Twitter’s
“OAuth”
protocol
required

minimal
application
development
time.
However,
it
resulted
in
substantial

complications
for
the
user’s
experience.
Requiring
each
user
to
login
using
their

personal
Twitter
user
account
not
only
increased
the
number
of
steps
for
access,

but,
more
importantly,
potentially
falsified
the
user’s
perception
of
safety.
An

important
feature
of
the
originally
designed
GeosocialFootprint
(2013)
was
to

showcase
the
public
nature
of
Twitter
data.
If
a
user
was
required
to
login
prior
to

viewing
their
footprint,
they
could
perceive
that
their
footprint
was
private.

The
use
of
a
Twitter
API
proxy
was
deemed
the
more
suitable
solution
for

GeosocialFootprint
(2013).
While
its
implementation
required
greater
development

time,
it
provided
for
the
user-‐friendly
experience
that
was
originally
intended.

Instead
of
requiring
the
user
to
login,
the
proxy
received
a
request
from
the

GeosocialFootprint
(2013)
user
and
subsequently
sent
the
same
request
on
behalf
of

the
user
to
the
official
Twitter
API.
The
proxy
to
authenticate
the
account
with

Twitter
then
used
a
Twitter
developer’s
token
and
key.
The
response
was
then

31

returned
to
the
original
user.
As
a
result,
the
token
and
key
were
registered
to

GeosocialFootprint
(2013)
and
stored
securely
by
the
proxy.

In
building
the
Twitter
API
proxy,
a
platform
as
a
service
(PAAS)
was
used
to

minimize
system
costs
and
administration
requirements.
Heroku
(2013)
was

selected
as
the
PAAS
provider
due
to
services
provided
at
no
cost
and
its
large
user

community.

3.3.2
User
Analytics

One
drawback
to
an
application
that
is
built
on
client-‐side
computing
logic
is
the

inability
to
store
user
metrics.
User
metrics
are
imperative
to
evaluate
and
validate

the
efficacy
of
an
application.
In
a
traditional
web
application,
the
same
database

that
stores
user
data
could
also
be
used
to
store
and
track
user
activity.
By
design,

GeosocialFootprint
(2013)
does
not
have
a
hosted
database
to
store
user

information
and
therefore
cannot
include
user
metrics.

Qbaka
(2013)
offered
an
adequate
solution
through
an
adaption
of
their
error

reporting
service
(2013).
GeosocialFootprint
(2013)
alters
the
standard
use
case
of

Qbaka
(2013)
by
reporting
client
actions
back
to
Qbaka’s
API,
instead
of
the

traditional
error
reports.
The
user
metrics
were
then
monitored
and
analyzed
using

Qbaka’s
(2013)
web
application.

32

Based
on
the
work
described
above,
the
next
chapter
discussed
the
finished

application
and
the
results
of
the
user
survey.

CHAPTHER
4:
RESULTS

4.1
Application
Function

By
design,
GeosocialFootprint
(2013)
provides
a
very
simple
user
interface.
The

home
page
of
GeosocialFootprint
(2013)
includes
a
summary
of
the
web
application,

instructions
for
use,
and
links
to
recent
news
coverage
this
web
application
has

received.
As
shown
in
Figure
2
the
instructions
are
in
bold
to
draw
attention.

Figure
2
-‐
GeosocialFootprint
(2013)
Landing
Page

33

A
user
views
a
GeosocialFootprint
(2013)
by
entering
the
desired
Twitter
username

in
the
input
box
and
clicking
the
button
titled
“Retrieve
Tweets”.

After
a
user

attempts
to
retrieve
tweets,
the
application
checks
the
validity
of
the
Twitter

username.
As
seen
in
figure
3,
if
the
username
is
valid
the
main
user
interface
box

updates
and
informs
the
user
that
the
application
is
retrieving
the
first
200
tweets.

If
the
Twitter
username
is
invalid
or
private
the
user
is
alerted
through
a
popup

window,
as
shown
in
Figures
4
and
5.

Figure
3
-‐
View
of
intermediate
web
page
after
a
valid
Twitter
username
has
been

entered

34

Figure
4
-‐
Popup
alerting
user
of
an
invalid
Twitter
username

After
a
valid
Twitter
username
has
been
entered
and
tweets
have
been
retrieved,

the
web
application
updates
the
main
user
interface
box
with
a
map.
Figure
4

represents
the
results
of
a
user
that
has
a
high-‐risk
rating
where
an
obvious

“footprint”
is
visible.
In
the
event
that
the
web
application
cannot
geo-‐locate
any

tweets
for
the
specified
user,
an
in
text
alert
is
added
between
the
input
box
and
the

map,
as
seen
in
Figure
6.

35

Figure
5
-‐
Map
results
of
a
high
risk
Twitter
user

Figure
6
-‐
Map
results
when
no
tweets
are
geo-‐located

36

4.1.1
User
risk,
alters,
and
suggestions

After
a
Twitter
username
has
been
validated,
tweets
have
been
retrieved,
and
a
map

has
been
generated
the
user
is
provided
with
dynamic
alerts,
a
location
over-‐sharing

risk
assessment,
and
suggestions
on
decreasing
said
risks.
Figure7
shows
the
three

columns
of
information
that
are
produced
for
the
user
after
a
successful
mapping
of

tweets.

Figure
7
-‐
Results
of
a
high-‐risk
user,
including
context
alerts
and
suggestions
to

reduce
risk

37

The
alerts
section
displays
the
results
of
the
context
filtering
analysis.
These
alerts

are
separated
into
three
categories
that
represent
the
likelihood
of
alert
accuracy:

low,
medium,
and
high.
This
accuracy
is
based
on
word
context
matching.
Key

phrases
were
built
off
of
previous
research
by
Humphreys
(2010)
and
were

designed
to
expose
locations
relating
to
a
users
home,
place
of
work,
and

educational
facility.
Key
phases
include
for
example,
"am
at
home"
which
is

generally
used
in
the
context
of
a
user
mentioning
they
are
at
their
place
of

residence.
When
the
occurrence
of
these
known
words
or
phrases
that
divulge

additional
location
information
increases
within
a
tweet,
so
does
the
accuracy

rating.
The
likelihood
category
is
not
only
displayed
as
text,
but
the
color
of
the
text

for
each
alert
changes
based
on
the
alert
level.
To
increase
the
usability
of
the
alerts

section,
when
the
user
clicks
on
an
alert,
the
map
centers
and
zooms
to
the
selected

alert.

The
risk
of
over-‐sharing
is
displayed
in
the
center
column.
The
risk
is
calculated

based
on
the
percentage
of
located
tweets
for
each
user.
As
the
percentage
of

located
tweets
increases,
so
does
the
risk.
The
risk
calculation
was
derived
from

findings
of
Weidemann
and
Swift
(2013)
which
breaks
down
the
observed
use
of

location
based
tweets
over
the
course
of
a
7
day
sampling
period.
Using
those

results
as
a
baseline,
the
following
risk
levels
were
created:
0%
is
no
risk,
between

3-‐6%
is
low
risk,
between
6-‐9%
is
average
risk,
between
9-‐12%
is
medium
risk,

38

between
12-‐24
percent
is
medium-‐high
risk,
and
above
24%
is
high
risk.
The

heading
of
the
risk
column
changes
according
to
the
resulting
risk
level.
This
text

also
changes
color
based
on
the
risk
level.
Below
the
risk
rating,
the
user
is
also

provided
with
a
summary
of
what
was
collected
for
the
specified
Twitter
user.

The
suggestions
column
dynamically
provides
the
user
with
suggestions
on
how
to

decrease
their
personal
over-‐sharing
risk.
This
suggestion
column
is
compiled
from

a
list
of
resources,
links,
and
videos
on
social
media
over-‐sharing(Twitter.com
2013,

Shelly
2013,
Groeneveld
et.
al.
2010).
The
web
application
analyzes
the
users

Twitter
trends
and
dynamically
selects
items
from
the
list
that
could
be
beneficial
to

the
specified
Twitter
username.

It
should
be
noted
that
no
guarantee
of
privacy
can
be
made,
even
after
following

the
steps
provided
in
suggestions.
Other
techniques
can
be
used
to
locate
tweets.

For
example
the
work
by
Weidemann
and
Swift
(2013)
examines
the
use
of

geocoding
to
locate
tweets
based
on
the
text
and
context
of
the
tweet.
While
turning

off
GPS
enabled
tweets
will
decrease
the
risk
of
location
profiling,
it
cannot
be

completely
negated
without
making
all
tweets
private.

Next,
the
user
is
automatically
given
the
opportunity
to
participate
in
a
user

feedback
survey,
as
described
in
the
following
section.

39

4.1.2
Anonymous
Survey

GeosocialFootprint
(2013)
has
an
internal
timer
that
asks
the
user
to
volunteer
for

an
anonymous
survey
two
minutes
after
tweets
are
retrieved
for
a
Twitter

username.
The
request
is
displayed
as
a
popup
on
the
main
page,
as
depicted
in

figure
8.
If
a
user
agrees
to
participate
in
the
survey,
they
are
forwarded
to
a
Google

Forms
page
where
they
are
asked
to
respond
to
the
questions
found
in
Figure
9.
A

summary
of
the
anonymous
survey
results
can
be
found
below
in
subsection
titled

“Summary
of
Survey
Results”,
while
all
results
collected
as
of
January
2014
can
be

found
in
Appendix
A.

Figure
8
-‐
Feedback
request
popup

40

Figure
9
-‐
Feedback
page

41

4.1.3
Data
Download

Upon
user
request,
data
download
functionality
was
added
to
GeosocialFootprint

(2013).
After
tweets
have
successfully
been
retrieved,
all
geo-‐located
tweets
are

available
for
download
as
a
comma-‐separated
values
(CSV)
file.
To
download
the

CSV
file,
a
user
must
successfully
plot
tweets
on
the
map
and
then
simply
click
the

“Download
Tweets”
button,
as
shown
in
Figure
10.
The
CSV
file
can
be
opened
in

most
modern
spreadsheet
applications,
along
with
numerous
GIS
software
suites.

The
CSV
file
provided
by
GeosocialFootprint
(2013)
contains
the
tweet
time,

latitude,
longitude,
and
text.
Example
contents
of
a
CSV
download
are
shown
in

Figure
11.

Figure
10
-‐
Data
download
feature
button

42

Figure
11
-‐
Contents
of
a
CSV
download
opened
in
Microsoft
Excel

While
the
data
download
may
have
many
uses,
it
was
specifically
designed
to
help

GIScience
users
more
fully
analyze
Twitter.com
data.
Due
to
the
limitations
in
the

Twitter
API,
the
Twitter
API
response
cannot
be
redistributed
in
its
original
form,

which
makes
sample
Twitter
data
hard
to
source
and
locate.
GeosocialFootprint

(2013)
removes
all
user
data,
as
required
by
the
Twitter
API
Terms
of
Service,
and

repackages
the
data
in
a
usable
format.
This
sample
data
can
then
be
used
by

GIScience
users
to
perform
more
advanced
analysis,
for
example
temporal
or

habitat
analysis.

43

4.2
Summary
of
Survey
Results

The
anonymous
user
survey
acts
as
tool
to
judge
the
efficacy
of
GeosocialFootprint

(2013).
The
participants
are
asked
to
rate
their
perceived
privacy
on
Twitter
prior

to
using
the
web
application,
then
again
after.
They
are
also
asked
to
rate
the

educational
value
of
the
suggestions
and
the
tool
as
a
whole.
Lastly,
users
are

provided
with
an
opportunity
to
provide
general
feedback
and
comments
through

manual
typing
into
a
form
field.

The
questions
asked
and
the
available
answers
are
as
follows:

1. Question:
Prior
to
using
GeosocialFootprint
(2013),
how
would
you
have

rated
your
location
privacy
on
Twitter.com?

Possible
Answer:
A
scale
from
one
to
ten
with
one
representing
“private”
and

ten
representing
“insecure”.

2. Question:
After
using
GeosocialFootprint
(2013),
how
do
you
rate
your

location
privacy
on
Twitter.com?

Possible
Answer:
A
scale
from
one
to
ten
with
one
representing
“private”
and

ten
representing
“insecure”.

3. Question:
How
would
you
rate
the
personalized
suggestions
you
received
on

reducing
your
over
sharing
risk?

44

Possible
Answer:
A
scale
from
one
to
ten
with
one
representing
“No
Help”
and

ten
representing
“Extremely
Helpful”.

4. Question:
Did
you
find
GeosocialFootprint
(2013)
educational?

Possible
Answer:
“Yes”,
“No”,
and
“Undecided”.

5. Question:
If
GeosocialFootprint
(2013)
located
areas
of
concern,
were
any
of

them
accurate?

Possible
Answer:
“Yes”,
“No”,
“Undecided”,
and
“GeosocialFootprint
(2013)

did
not
locate
areas
of
concern
for
my
account”.

6. General
Comments
section:
“We're
interested
to
hear
what
you
have
to
say.

Please
share
any
general
comments,
concerns,
or
feature
requests.”

Possible
Answer:
Blank
text
box.

Figure
12
shows
the
summarized
results
of
question
one.

Only
16%
of
the

anonymous
users
reported
an
insecurity
rating
of
8
or
higher
with
respect
to
their

feelings
about
using
Twitter,
with
10
being
the
most
insecure,
while
50
percent
felt

very
confident
in
their
privacy
with
a
rating
of
3
or
less.

45

Figure
12
-‐
Summarized
responses
to
question
one
on
the
user
survey

Figure
13
represents
the
summarized
results
of
anonymous
user’s
perceived

privacy
after
using
GeosocialFootprint
(2013).
The
results
from
question
2.
show

that
roughly
half
(49%)
of
users
who
responded
to
the
survey
indicate
that
they

have
strong
confidence
in
their
privacy
after
using
the
application.
However
users

that
recorded
an
insecure
rating
of
8,
9,
or
10
double
compared
to
question
1.

accounting
for
30%
of
the
users.
The
results
from
question
2.show
that

GeosocialFootprint
(2013)
helped
increase
awareness
of
social
media
over-‐sharing

insecurities.
In
addition,
the
number
of
users
that
rated
their
location
privacy
a
1,

meaning
most
private,
increased
from
22%
for
question
1.to
30%
for
question
2.

This
shows
that
the
tool
not
only
highlights
insecurities
but
it
was
also
able
to
instill

confidence
in
some
users
about
their
privacy.

46

Figure
13
-‐
Summarized
responses
to
question
two
on
the
user
survey

Figure
14
summarizes
the
results
from
question
three.
The
bimodal
distribution
of

the
graph
shows
that
users
either
found
the
personalized
suggestions
extremely

helpful
or
no
help
at
all.
Many
users
that
reported
“no
help”
also
left
comments

suggesting
their
Twitter
username
didn’t
produce
any
geo-‐located
Tweets,
therefore

no
personalized
suggestions
were
provided.

47

Figure
14
-‐
Summarized
responses
to
question
three
on
the
user
survey

Figure
15
summarizes
the
results
from
questions
four
and
five.
Question
4.

responses
provide
the
most
convincing
results
in
favor
of
GeosocialFootprint

(2013)s
effectiveness
in
educating
social
media
users
of
location
privacy.
An

overwhelming
82%
found
GeosocialFootprint
(2013)
to
be
educational,
with
only

8%
not
finding
it
educational,
while
an
additional
10%
of
users
remained

undecided.

Results
for
question
five,
summarized
as
well
in
figure
15,
show
that
54%
of
the

users
did
not
have
areas
of
concern
highlighted
for
their
retrieved
tweets.
Of
the

remaining
users,
over
half
found
that
the
areas
of
concern
that
were
highlighted

48

were
accurate.
This
is
strong
evidence
that
word
context
matching
is
indeed
a

viable
solution
for
collected
meta
data
about
location
enabled
tweets.

Figure
15
-‐
Summarized
responses
to
questions
four
and
five
on
the
user
survey

Question
6
asks
the
user
for
generic
feedback
on
the
application
through
a
text
box.

The
responses
provided
in
Question
6
personally
validate
the
efficacy
of
the
web

application.
One
user
commented
that
they
made
their
"tweets
private
after"

viewing
GeosocialFootprint
(2013).
They
continue
by
saying
"I
was
high
risk
with

148
out
of
199"
being
located
"and
you
knew
where
I
worked
and
lived".
This
user's

comment
shows
that
at
least
in
one
case
GeosocialFootprint
(2013)
was
able
to

educate
and
inform
a
user

not
only
about
their
personal
location
privacy
but
also

how
to
proceed
with
reducing
their
over-‐sharing
risks.
Another
user
had
similar

results
and
stated:
"
Nice
article.
This
is
kind
of
scary.
While
I
personally
don't
have

49

any
locations
on
the
map,
my
kids
each
have
our
home
and
their
local
hangouts

visible.
Thanks
for
helping
me
be
a
better
parent."
This
comment
is
another

testament
to
the
efficacy
of
GeosocialFootprint
(2013)
as
it
empowered
a
parent
to

help
product
their
children.

Of
course
not
all
the
comments
were
positive.
GeosocialFootprint
(2013)
was
met

with
criticism
as
well.
One
user
stated:
"I
consider
it
important,
useful,
and
good

etiquette
to
share
my
location
in
my
tweets".
This
comment
reaffirms
the
notion

that
the
concept
over-‐sharing
is
still
personally
subjective.

These
combined
results
provide
evidence
that
GeosocialFootprint
(2013)
was

successful
in
completing
its
main
goal:
educating
and
informing
social
media
users

of
the
privacy
risks
of
location
over-‐sharing.
In
addition
the
results
conclude
that

word
context
matching
can
be
implemented
in
social
media
feeds
to
gleam

additional
location
information
from
geo-‐enabled
tweets.
Finally
the
personalized

suggestions
on
how
to
increase
social
media
privacy
were
successful
when
users

have
geo-‐enabled
tweets.

50

CHAPTER
5:
Conclusion

5.1
Main
Advances

GeosocialFootprint
(2013)
is
unique,
unlike
any
other
web
application.
No
other

online
mapping
application
publically
available
allows
Twitter
user
to
visualize
their

tweets
as
an
area
of
interest,
provides
ambient
location
data
alerts,
performs
over-‐
sharing
risk
assessments,
or
educates
the
user
through
dynamic
and
customized

reporting.

GeosocialFootprint
(2013)
is
like
all
other
innovations
in
that
it
builds
upon
the

successes
of
others.
By
combining
historically
successful
geosocial
tools
into
one

geosocial
web
application,
GeosocialFootprint
(2013)
is
able
to
push
the
boundaries

of
geosocial
location
privacy
and
user
education.

Beyond
the
principles
of
educating
and
informing
social
media
users
of
the
risks
of

overs-‐haring,
GeosocialFootprint
(2013)
also
expands
upon
the
body
of
knowledge

by
advancing
the
capabilities
of
client
side
web
application
processing
for
geospatial

tools.
Traditionally
users
equate
geospatial
web
tools
with
expensive
and
dedicated

enterprise
servers.
Technology
today
allows
for
the
large
majority
of
geospatial

processing
to
occur
in
the
web
browser
on
the
client’s
(user’s)
machine.
As
a
proof

of
concept,
GeosocialFootprint
(2013)
greatly
reduces
the
entry-‐level
barrier
for

future
geospatial
tools.

51

5.2
Next
Steps

Advances
in
machine
learning,
data
analytics,
temporal
analysis,
and
entity

recognition
provide
for
further
research
in
the
area
of
geosocial
location
privacy.

Research
by
Weidemann
and
Swift
(2013)
shows
that
geocoding
entire
tweets

through
entity
recognition
and
accuracy
assessments
can
increase
the
breadth
of

geo
data
from
3%
of
the
Twitter
API
fire
hose
stream
to
nearly
30%.

GeosocialFootprint
(2013)
proves
this
by
providing
simple
metrics
on
user
twitter

habits
that
helps
to
inform
the
user.
Delivering
further
data
analytics
could
improve

the
efficacy
of
the
user’s
education.
The
addition
of
temporal
analysis
may
also

increase
the
accuracy
of
the
area
of
concern
alerts.
While
the
risk
assessments
and

alert
levels
were
deemed
informative
through
the
anonymous
survey,
using
basic

machine
learning
could
expand
upon
this.

Geocoding
of
tweets
is
not
a
new
subject.
It
has
been
researched
by
Huck
et
al.

(2012),
who
found
that
while
limitations
still
exist
in
current
methodologies,
viable

location
data
can
be
processed
using
a
multiple
pass
geocoding
technique.

Weidemann
and
Swift
(2013)
also
found
that
it
is
not
only
plausible
but
it
is
also

being
used
as
a
intelligence
gathering
tool
by
governments,
criminals,
and
even

commercial
entities.
Unfortunately
the
general
public
still
does
not
have
access
to
a

tool
that
can
help
them
visualize
the
ambient
geosocial
data
they
also
disclose
in

their
tweets.

52

A
new
geocoding
tool
is
planned
for
a
second
version
of
GeosocialFootprint
(2013),

based
off
the
research
by
Weidemann
and
Swift
(2013).
It
will
further
develop
this

open
source
application
to
allow
for
geographic
entity
recognition
and
geocoding.

Entity
recognition
extracts
usable
information
from
strings
of
text.
That
usable

information
can
then
be
compared
against
a
gazetteer
and
geographic
coordinates

can
be
assigned
to
the
string
of
text
as
a
whole.
Geocoding
would
increase
the

number
of
geographic
coordinates
visualized
on
the
map
and
ultimately
help
better

shape
a
user’s
GeosocialFootprint
(2013).
Geocoding
functionality
was
not
included

in
the
scope
of
this
project
due
to
depreciated
functionality
of
the
Google
Maps
API,

as
used
in
Weidemann
and
Swift
(2013).
Alternatives
to
entity
recognition
and

natural
language
processing
have
been
investigated
and
will
be
built
into
the

application
for
a
future
release.

Beyond
geocoding,
GeosocialFootprint
(2013)
could
be
improved
by
increasing
the

analytics
performed
on
the
users
tweets.
The
results
of
the
survey,
as
discussed
in

chapter
4,
show
that
users
found
the
simple
data
analytics
informative.
It
can
also
be

derived
from
user
survey
responses
that
increasing
the
data
metrics
and
analytics

would
also
increase
users’a
bility
to
become
more
educated
about
their
privacy

through
the
web
application.

53

Not
only
would
general
data
analytics
improve
the
efficacy
of
the
web
application

but
temporal
analysis
of
the
tweets
may
also
improve
the
accuracy
of
alerts
and

suggestions.
Nagarajan
et
al.
(2009)
points
out
that
the
time
of
the
day
a
tweet
is

published
can
be
a
strong
indicator
for
assumed
location
of
the
user.
In
the
context

of
GeosocialFootprint
(2013)
this
can
be
used
to
more
accurately
define
areas
of

concern
for
the
user.
With
the
current
alerts,
word
context
matching
is
singularly

used
to
perform
analytics.
Including
temporal
analysis
may
improve
such
findings

by
detecting
patterns
in
temporal
fluctuations
and
correlating
those
to
anticipate
or

predict
location.

Machine
learning
could
improve
upon
the
analytics
even
further
by
refining
the
data

processing
and
analytical
methods
based
on
the
continued
use
of
the
application

and
the
results
found
therein.
For
example,
the
risk
assessment
method
could

dynamically
respond
to
the
continued
use
of
the
web
application
and
adjust
the
risk

classifications
based
on
statistical
findings
such
as
a
decrease
in
the
mean

percentage
of
geo
enabled
tweets.
Alternatively
entity
recognition
and
gazetteers

could
be
modified
in
the
geocoding
process
as
users
verify
the
accuracy
of
such

findings
through
a
user
survey.

54

5.3
The
Future
of
GIS
and
social
media

Technology
changes
faster
than
most
users
can
keep
up.
In
the
context
of
social

media
the
change
has
brought
on
many
privacy
concerns,
a
small
part
of
which
of

which
GeosocialFootprint
(2013)successfully
addresses.
While
sometimes
this
rapid

change
brings
on
negative
consequences,
many
times
technology
evolves
to
improve

society.
Social
media
has
seen
negative
change
in
the
past
yet
is
now
entering
a
new

era
of
improvement.

Social
media
providers,
including
Twitter,
have
made
initiatives
to
limit
accessibility

to
user
data
through
the
use
of
authenticated
APIs.
Other
social
media
providers

have
tried
to
better
manage
data
sharing
options
and
security.
Through
trial
and

error,
social
media
providers
will
find
a
solution
that
enables
rich
immersive
social

media
content
without
sparing
the
user’s
privacy.

Applications
that
expand
new
technology
and
highlight
weakness
are
necessary
in

the
overall
balance
of
social
media
privacy.
Location
privacy
specifically
will
see
a

slew
of
applications
that
either
harnesses
the
locational
power
of
social
media
to

provide
innovative
uses,
or
on
the
flip
side,
applications
that
exploit
the
weaknesses

of
such
network.

55

Of
particular
emphasis
will
be
direct
and
indirect
VGI.
With
indirect
VGI
the
user

anonymously
and
autonomously
contributes
to
a
geographic
information
collection

through
active
application
and
sensors.
An
example
of
this
is
WAZE,
a
traffic

application
that
becomes
more
powerful
as
more
users
contribute
to
the
geospatial

system(Smith
2011).

OpenStreetMap
is
another
example
of
direct
VGI
[ref].
Users

directly
contribute
to
an
obvious
geospatial
system.
While
neither
of
these
VGI

examples
use
Twitter
as
their
communication
network,
the
same
information
can
be

transferred
through
social
media
protocols.
Using
3
rd

party
API’s,
such
as
Twitter,
to

communicate
VGI
can
greatly
decrease
the
development
time
and
effort
required
to

generate
new
similar
applications
and
thus
enable
more
successful

implementations.

Indirect
VGI’s
that
use
social
media
as
a
sensor
are
already
in
use
today.
The
Federal

Emergency
Management
Agency
(FEMA)
uses
social
media
event
tracking
as
an

indirect
VGI
to
inform
decision
makers
in
the
event
of
an
emergency(Fugate
2011).

Twitter
users
do
no
actively
communicate
on
Twitter
for
the
purpose
of
informing

FEMA’s
VGI,
but
instead
the
FEMA
VGI
actively
listens
for
geospatial
content
related

to
their
interests.
This
same
concept
will
probably
be
applied
to
different
mediums

in
the
future.
Effective
privacy
constraints
need
to
be
in
place
to
reduce
the
risk
of

nefarious
VGI’s
while
still
empowering
well-‐meaning
entities
like
FEMA
to
save

lives.

56

The
future
not
only
holds
a
lot
of
potential
opportunities
to
advance
the
use
of

geosocial
data
but
to
also
educate
and
inform
the
general
public
more
about
their

social
media
location
privacy.
As
the
results
of
this
thesis
suggest,
simple

visualization
is
an
effective
tool
to
educate
users.
It
is
the
opinion
of
the
author
that

social
media
outlets
should
perform
user
outreach
and
allow
users
to
proactively

monitor
their
social
media
footprint
within
their
existing
web
application

frameworks.
This
is
the
responsible
path
forward,
for
all
concerned.

57

References

Ahn,
G.-‐J.,
M.
Shehab
&
A.
Squicciarini
(2011)
Security
and
privacy
in
social

networks.
Internet
Computing,
IEEE,
15,
10-‐12.

Asur,
S.
&
B.
A.
Huberman
(2010)
Predicting
the
future
with
social
media.
arXiv

preprint
arXiv:1003.5699.

Barkhuus,
L.
(2004).
Privacy
in
location-‐based
services,
concern
vs.
coolness.
In

Proceedings
of
workshop
paper
in
mobile
HCI
2004
workshop:
location

system
privacy
and
control.
Glasgow,
UK.

Barkhuus,
L.
&
A.
Dey.
(2003).
Location-‐based
services
for
mobile
telephony:
a
study

of
users’
privacy
concerns.
In
Proc.
Interact,
709-‐712.
Citeseer.

Barr,
J.,
Tetlaw,
A.,
&
Simoneau,
L.
(2010).
Host
your
web
site
in
the
cloud:
Amazon

web
services
made
easy.
SitePoint.

Bent
Design
Studios
(2012).
Developer
Guide
(HTML,
CSS,
CSS3,
LESS)
V
2.0.

Retrieved
at:
http://www.bentdesignstudio.com/v2/2012/03/twitter-‐
bootstrap-‐2-‐photoshop-‐template-‐psd/

Beresford,
A.
R.
&
F.
Stajano
(2003)
Location
privacy
in
pervasive
computing.

Pervasive
Computing,
IEEE,
2,
46-‐55.

Bettini,
C.,
X.
Wang
&
S.
Jajodia
(2005)
Protecting
privacy
against
location-‐based

personal
identification.
Secure
Data
Management,
185-‐199.

"Brazil
Police
Bust
Kidnappers
Who
Browsed
Social-‐networking
Sites
for
Victims."

Fox
News.
FOX
News
Network,
02
Aug.
2010.
Web.
02
Oct.
2013.

Bootstrap
(2013)
GetBootstrap.com.
Retreived
at:
http://getbootstrap.com/

Colbert,
M.
(2001).
A
diary
study
of
rendezvousing:
implications
for
position-‐aware

computing
and
communications
for
the
general
public.
In
Proceedings
of
the

2001
International
ACM
SIGGROUP
Conference
on
Supporting
Group
Work,

15-‐23.
ACM.

Danezis,
G.,
S.
Lewis
&
R.
Anderson.
(2005).
How
much
is
location
privacy
worth.
In

Fourth
Workshop
on
the
Economics
of
Information
Security.
Citeseer.

58

Duckham,
M.
&
L.
Kulik
(2006)
Location
privacy
and
location-‐aware
computing.

Dynamic
&
Mobile
GIS:
Investigating
Change
in
Space
and
Time,
34-‐51.

ESRI
(2013)
ArcGIS
API
for
JavaScript,
Get
Started.
Retrieved
at:

https://developers.arcgis.com/javascript/

Ferrari,
L.,
A.
Rosi,
M.
Mamei
&
F.
Zambonelli.
(2011).
Extracting
urban
patterns

from
location-‐based
social
networks.
In
Proceedings
of
the
3rd
ACM

SIGSPATIAL
International
Workshop
on
Location-‐Based
Social
Networks,
11.

ACM.

Friedland,
G.
&
R.
Sommer.
(2010).
Cybercasing
the
joint:
On
the
privacy

implications
of
geo-‐tagging.
In
Proc.
USENIX
Workshop
on
Hot
Topics
in

Security.

Fugate,
C.
(2011).
Understanding
the
Power
of
Social
Media
as
a
Communication

Tool
in
the
Aftermath
of
Disasters.
112th
Cong.

Gnip
(2013)
Gnip
Support:
APIs.
Retrieved
at
http://support.gnip.com/apis/

Goodchild,
Michael
F.
(2007).
Citizens
as
sensors:
the
world
of
volunteered

geography.
Geojournal,
69(4),
211-‐221.

Groeneveld,
F.,
Borsboom,
B.,
Amstel,
B.
(2010),
Center
for
Democracy
&

Technology:
Over-‐sharing
and
Location
Awareness.
Retrieved
at:

https://www.cdt.org/blogs/cdt/over-‐sharing-‐and-‐location-‐awareness

Heroku
(2013)
Heroku.com
[Web
Application
Software],
Retreived
at:

https://www.heroku.com/

Huck,
J.,
Whyatt,
D.,
&
Coulton,
P.
(2012).
Challenges
in
Geocoding
Socially-‐
Generated
Data.

Humphreys,
L.,
P.
Gill
&
B.
Krishnamurthy.
(2010).
How
much
is
too
much?
Privacy

issues
on
Twitter.
In
Conference
of
International
Communication
Association,

Singapore.

Iachello,
G.,
I.
Smith,
S.
Consolvo,
G.
Abowd,
J.
Hughes,
J.
Howard,
F.
Potter,
J.
Scott,
T.

Sohn
&
J.
Hightower
(2005)
Control,
deception,
and
communication:

Evaluating
the
deployment
of
a
location-‐enhanced
messaging
service.

UbiComp
2005:
Ubiquitous
Computing,
903-‐903.

jQuery
(2013)
jquery.com/.
Retrieved
at:
http://jquery.com/

59

Kaasinen,
E.
(2003)
User
needs
for
location-‐aware
mobile
services.
Personal
and

ubiquitous
computing,
7,
70-‐79.

Krumm,
J.
(2009)
A
survey
of
computational
location
privacy.
Personal
Ubiquitous

Comput.,
13,
391-‐399.

Leaflet
(2013)
An
Open-‐Source
JavaScript
Library
for
Mobile-‐Friendly
Interactive

Maps.
Retrieved
at:
http://leafletjs.com/reference.html

Leetaru,
K.,
S.
Wang,
G.
Cao,
A.
Padmanabhan
&
E.
Shook.
(2013).
Mapping
the
global

Twitter
heartbeat:
The
geography
of
Twitter.

Leetaru,
K.
H.
(October
2012)
Fulltext
Geocoding
Versus
Spatial
Metadata
for
Large

Text
Archives:
Towards
a
Geographically
Enriched
Wikipedia.
D-‐Lib

Magazine,
18.

Madden,
M.
(2012)
Privacy
management
on
social
media
sites.
Pew
Internet
Report.

Mao,
H.,
X.
Shuai
&
A.
Kapadia.
(2011).
Loose
tweets:
an
analysis
of
privacy
leaks
on

twitter.
In
Proceedings
of
the
10th
annual
ACM
workshop
on
Privacy
in
the

electronic
society,
1-‐12.
Chicago,
Illinois,
USA:
ACM.

Microsoft
MSDN
(2013)
Web
Development
.
Retreived
at:

http://msdn.microsoft.com/en-‐us/library/aa155073.aspx

Paddock,
R,
Petersen,
J.
(2012)
Practical
jQuery.
Apress

PleaseRobMe(2012)
Please
Rob
Me,
Raising
awareness
about
over-‐sharing.
[Web

Application
Software].
Retrieved
at:
http://www.Pleaserobme.com

Qbaka
(2013)
Qbaka.com
[Web
Application
Software].
Retrieved
at:

https://qbaka.com/

Ruiz
Vicente,
C.,
D.
Freni,
C.
Bettini
&
C.
S.
Jensen
(2011)
Location-‐related
privacy
in

geo-‐social
networks.
Internet
Computing,
IEEE,
15,
20-‐27.

Ready
or
Not
(2013)
Teaching
Privacy,Ready
or
Not
[Web
Application
Software].

Retrieved
at:
http://teachingprivacy.icsi.berkeley.edu:8080/

Shelly
K
(2013)
USA
Today:
Social
Media
makes
it
easy
to
share
and
overshare
alike.

Retrieved
at:

60

http://www.usatoday.com/story/news/nation/2013/05/09/social-‐media-‐
oversharing/2148941/

Stefanidis,
A.,
A.
Crooks
&
J.
Radzikowski
(2012)
Harvesting
ambient
geospatial

information
from
social
media
feeds.
Geojournal,
1-‐20.

Smith,
C.
E.
(2011).
Geospatial
encountering:
Opportunistic
information
discovery
in

web-‐based
GIS
environments.
Proceedings
of
the
American
Society
for

Information
Science
and
Technology,
48(1),
1-‐4.

Sublime
Text
2(2012)
SublimeText.com
[Desktop
Application
Software].
Retrieved

at:
http://www.sublimetext.com/2

Trello
(2013)
Trello.com
[Web
Application
Software]
Retrieved
at:

http://www.trello.com/

Twitter
(2013)
Twitter.com,
Help
Center:
Adding
your
location
to
a
Tweet.

Retrieved
at:
https://support.twitter.com/articles/122236-‐how-‐to-‐tweet-‐
with-‐your-‐location

Tweetping
(2012)
Tweetping.net
[Web
Application
Software]
Retrieved
at:

http://tweetping.net/

Tweography
(2011)
Tweography:
Your
tweets
on
a
map.
[Web
Application

Software]
Retrieved
at:
http://www.tweography.com

UMLet
(2013)
Umlet.com
[Web
Application
Software]
Retrieved
at:

http://www.umlet.com/

University
of
California,
San
fransico
IT
(2013):

Software
Development
Lifecycle

(SDLC).
Retrieved
at:

https://it.ucsf.edu/pages/software-‐development-‐
lifecycle-‐sdlc

Wang,
L.,
Von
Laszewski,
G.,
Younge,
A.,
He,
X.,
Kunze,
M.,
Tao,
J.,
&
Fu,
C.
(2010).

Cloud
computing:
a
perspective
study.
New
Generation
Computing,28(2),

137-‐146.

WeKnowYourHouse(2012)We
Know
Your
House,
Another
Social
Networking

Privacy
Experiment.
[Web
Application
Software].
Retrieved
at:

http://www.weknowyourhouse.com

61

We
Know
What
You’re
Doing
(2011)
We
Know
What
You’re
Doing,
a
social

networking
privacy
experiment[Web
Application
Software].
Retrieved
at:

http://weknowwhatyouredoing.com/

62

Appendix
A:
Survey
Results

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
8/27/2013
16:07:31
1 1 6 Yes Undecided
8/28/2013
10:18:51
7 6 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
interesting app will be watching to see how it develops
8/28/2013
13:18:00
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

8/28/2013
14:18:20
1 1 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great app for quickly showing people how much information they may
be strewing around the Internet without even knowing. Keep up the
good work.
8/28/2013
17:01:52
3 8 8 Yes Yes I will stop sharing my location so often. It knew where I work.
8/29/2013
5:56:26
10 9 7 Yes No Nothing new under the sun here.
Everyone and their mother know (or should know) that the core
business of social based media-companies is
logging/tracking/directing of content/places, with the intent of selling
this information to 3rd party. Whether that be for advertising or other.
8/29/2013 4 8 7 Yes Yes I will check back for the new features. You should have an opt-in e-

63

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
12:02:24 mail list to keep people up to date.
8/30/2013
10:02:25
6 9 8 Yes Undecided
8/30/2013
10:02:25
6 9 8 Yes Undecided
8/30/2013
16:26:18
3 2 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/1/2013
17:41:06
3 8 10 Yes Yes
9/1/2013
20:43:56
3 7 9 Yes No
9/2/2013
5:27:47
5 1 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/2/2013
13:35:24
7 7 4 Yes No As a Geography student I found this tool very interesting. Nice work!
9/3/2013
14:32:29
7 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Fortunately, I'm very conscious of the geolocation information I share.
(I have given talks in this area to teens before.)

I was actually surprised that only 1/200 was flagged. But because I
tweet so much, the last 200 tweets may not be a good sampling of
"activity" tweets. Thus, my location privacy rating was (pleasantly)
much lower than I expected, and I did not receive personalized
suggestions on reducing oversharing risk.
I would love to see Instagram, Foursquare and the like eventually
incorporated into this app. Great tool!

64

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/3/2013
18:30:17
3 1 6 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

9/3/2013
21:54:49
6 9 8 Yes Yes
9/3/2013
21:58:08
4 2 7 Yes Yes
9/3/2013
21:59:17
3 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/3/2013
22:00:17
1 1 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

9/3/2013
22:01:09
4 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Very interesting! And useful.

Re: "How would you rate the personalized suggestions you received
on reducing your over sharing risk? ": There were no suggestions
because I showed no risk.
9/3/2013
22:06:57
7 10 6 Undecided No
9/3/2013
22:07:10
1 1 10 Yes Geosocial
Footprint did
not locate
areas of

65

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/3/2013
22:15:20
5 5 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account
This website does not work. I enabled all scripts and nothing
happened when I pushed the various buttons
9/3/2013
22:27:21
5 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/3/2013
22:59:45
7 6 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
0:37:13
3 2 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
No risk in my case may be due to language not being english
9/4/2013
1:12:59
3 2 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
2:59:56
5 2 10 Yes Geosocial
Footprint did
not locate
areas of
I am strongly interested in Internet privacy and appreciate anything
that helps me manage my online presence, particularly location data
so I hope you get this up and running. Good luck.

66

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
4:52:38
8 9 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
5:30:58
8 9 10 Yes Yes Scary, but informative!
9/4/2013
5:59:06
3 1 5 Yes Geosocial
Footprint did
not
Great service, keep up the good work
9/4/2013
7:08:57
1 1 5 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
Interesting. I came out clean but when I looked at my tweets they do
have locational references, fireflies, bluebirds, bears, forsythia, snow
storm, visiting NYC... vague but locational
9/4/2013
7:19:22
1 1 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great idea - finally a website that seeks to educate about geotagging
rather than use the info for other means. Nice work.
9/4/2013
7:26:23
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
7:26:36
2 2 3 Yes Geosocial
Footprint did
not locate
areas of
concern for
Great idea, well implemented.

67

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
7:38:43
1 1 7 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
My account was just fine but I checked some friends and Geosocial
Footprint successfully told me where my friend lived (I knew anyway).
Scary accurate!
9/4/2013
7:49:17
1 1 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
8:14:32
3 3 7 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
8:33:46
1 10 8 Yes Yes
9/4/2013
8:36:59
3 7 10 Yes Yes
9/4/2013
8:37:55
3 3 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I don't have any of the location features turned on. What would be
interesting is trying to determine my location based on people I follow
and who I interact with.

Really interesting project. Nice work
9/4/2013
8:38:06
2 2 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

68

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
8:48:00
3 1 5 Yes No
9/4/2013
8:48:07
8 8 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great work. Extremely useful for people who arent asare of the
danger of geotags
9/4/2013
9:05:39
1 9 8 Yes Yes
9/4/2013
9:57:13
2 3 2 Yes No Geosocial confirmed that I was being safe with my geolocation and
not revealing my home or work areas.
9/4/2013
9:59:25
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account
None of the accounts I tested showed any location concerns at all,
which makes me wonder if it's really working.
9/4/2013
10:27:40
3 2 6 Yes Geosocial
Footprint did
not locate
areas of
concern for

9/4/2013
11:06:21
9 7 1 No Yes I only geotag tweets from places I don't care about people knowing I
am. You didn't pick up any private places
9/4/2013
11:25:21
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
You should be clear about your privacy policy and user agreement.

Also- if this is meant to be educational, in the sense that it helps users
become more aware of risks associated with tweets/locations- you
should consider getting a subject matter expert on personal protection
to provide guidance and advice- thereby helping give perspective.
9/4/2013
11:31:31
7 10 5 Yes Yes

69

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
11:35:20
4 8 9 Yes Yes
9/4/2013
12:10:44
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Going 200 tweets is not enough
9/4/2013
12:22:27
6 7 8 Yes Yes
9/4/2013
12:42:36
8 2 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
13:12:33
3 4 6 Undecided No
9/4/2013
13:15:35
2 2 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
13:39:02
6 8 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
13:53:15
1 1 4 Yes Geosocial
Footprint did
not locate
areas of
Just started exploring the website, but I love the idea. keep up the
good work!

70

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
13:55:08
7 2 8 Yes Yes
9/4/2013
14:06:45
1 1 5 Undecided No It gave my location as somewhere in the US, wheras I am in the UK
9/4/2013
14:32:09
3 2 7 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Very neat tool. Thank you very much.
9/4/2013
14:42:28
4 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Good Luck with the research. It did not find any location identifiers for
my account.

I have a twitter acct but rarely tweet. I mainly lurk in background to
gather info, insigtht, and ideas in my own PLN.

Overall for all social media, I try to be careful regarding security
threats, identity, and location.
9/4/2013
15:19:08
3 3 3 Yes No
9/4/2013
15:43:18
4 5 10 Yes Undecided
9/4/2013
15:44:32
2 1 8 Yes Geosocial
Footprint did
not locate
areas of
concern for

9/4/2013
16:10:45
8 7 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great idea, great program!

71

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
17:15:31
8 8 8 Yes Yes
9/4/2013
17:19:38
2 2 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
17:31:12
1 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
17:38:27
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account
Ui isn't intuitive controls are where the footer normally resides. Most
of the page is taken up telling you how twitter works nothing about
how to make it work. It told me it was loading 200 tweets eventually
after i'd finally figured it out then showed at map of America
somewhere I've never been. It didn't say it had found tweets or not.
So maybe i haven't figured out how it works. The colour of the buttons
appears to be significant but its not clear why or how. If I clear tweets
am I removing them from the analysis or from twitter? Fairly poor user
experience
9/4/2013
17:57:20
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
There are a couple of typos in the intro text on the home page:

This website provides twitter users with an oppurunity to view their
geosocial footprint. In additional it informs users of some potential
areas of concern with their current sharing habits. To begin, enter you

oppurunity -> opportunity
additional -> addition,
enter you -> enter your
9/4/2013
19:25:59
8 8 8 Yes Yes

72

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
19:29:56
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
19:45:11
2 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
GQ through Reddit brought me here
9/4/2013
19:48:50
4 3 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
This is an excellent educational tool!
9/4/2013
20:15:49
3 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
20:21:19
3 1 2 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
20:26:43
3 2 1 Yes Geosocial
Footprint did
not

73

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
20:56:34
3 3 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
20:59:29
3 2 10 Yes Yes Great program - congrats!
9/4/2013
21:05:16
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
21:27:48
1 9 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I was reading The Hill Newspapers signed in to Twitter using
Google. Newsmax asked to vote in a poll. I did complete the poll
than Newsmax
pulled up a map and asked me is this where you live. I was my
home.
I have never had this happen to me before it frightened me. How do I
get off of Google information what-ever. Google shares too much
personal information. I recently signed up to sign in with Google, I
was with Yahoo, no problem with Yahoo. I didn't use my name with
yahoo, but Google used my name. I will delete the account.
9/4/2013
22:21:22
1 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
22:33:32
2 1 10 Yes Geosocial
Footprint did
not locate
areas of

74

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/4/2013
22:43:06
3 1 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/4/2013
23:07:06
1 1 2 Yes Yes
9/4/2013
23:13:55
4 7 7 Yes Yes
9/5/2013
0:29:01
2 1 3 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
1:00:42
10 1 10 Yes Yes
9/5/2013
1:42:20
5 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I appreciate the thought and effort that went into producing this site. In
general, I want to be more informed about my privacy in relation to my
web use.
9/5/2013
2:18:31
8 8 6 Yes Yes
9/5/2013
3:43:00
1 1 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great idea! Very useful and important work. Thank you :)
9/5/2013
5:03:06
3 4 1 Yes Geosocial
Footprint did
So far it seems to be stymied by private accounts, which is good. But
this raises the question, if a profile isn't public, who could do this sort

75

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/5/2013
6:10:58
7 4 6 Yes Yes
9/5/2013
6:15:18
10 10 5 No No
9/5/2013
6:26:22
3 1 3 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great idea, I'm sure a lot of work went into it. Thanks
9/5/2013
6:28:56
8 5 6 Yes Yes Very interesting project, thanks.
9/5/2013
6:32:04
3 3 5 Yes Yes
9/5/2013
6:33:01
4 6 7 No Geosocial
Footprint did
not locate
areas of
concern for
my account
Why the fuck would anybody wsnt to know where I am? I'm no crook,
or runaway or anything that someone needed to know my location. I
live a quiet life and just want to be left alone. My life is mediocre and I
have no idea why someone would think my activities are even worth
tracking.
9/5/2013
6:43:12
7 4 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
7:40:53
3 10 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
7:48:40
3 8 5 Yes Geosocial
Footprint did

76

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/5/2013
8:09:52
5 5 9 Yes Yes
9/5/2013
8:36:21
1 1 10 Yes Yes This is awesome! I teach kids about social media and the data they
leave on the internet, this is going to be a great tool!
9/5/2013
10:12:50
6 9 10 Yes Yes
9/5/2013
10:14:54
1 1 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
This is bloody brilliant. I'm always concerned about making sure my
geocoding isn't activated on my profiles for just about everything. And
I'm glad to see that twitter or my phone haven't accidentally slipped
that into my tweets.
9/5/2013
10:27:39
1 1 9 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
10:49:40
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great tool!
9/5/2013
11:07:45
2 2 3 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
thanks for putting the app up
9/5/2013
11:10:56
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
Really interesting project, I want to know a lot more.

77

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/5/2013
11:26:00
1 6 8 Yes Yes Nice job. Will raise awareness about oversharing.
9/5/2013
11:29:20
5 7 8 Yes Yes
9/5/2013
11:30:48
4 1 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
This is a great tool. Immediately I did realize the power for it to be
used to find info on others. I suggest an OAuth to use it. But more
people are starting to question Internet safety. The hidden geolocate
information concerned me most. Also what if followers/following also
help pinpoint? Scan
9/5/2013
11:32:50
8 9 6 Yes Yes I knew it but it's good to be reminded!
9/5/2013
11:40:01
3 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
12:13:39
7 7 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
13:10:56
2 9 8 Yes Undecided
9/5/2013
13:11:57
3 8 8 Yes Undecided
9/5/2013
13:22:30
7 10 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
It's a good eyeopener.

78

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/5/2013
13:53:39
5 5 3 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
The suggestions should be a bit more specific. Perhaps detailing what
apps that are posting the geolocations would be best. in my case, my
geotagged tweets are postings from foursquare which i intentionally
shared on Twitter.
9/5/2013
14:51:35
10 1 4 Yes Yes
9/5/2013
15:59:57
4 3 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Surprised, but I don't tweet from a mobile phone. :) I think that makes
a difference?
No Risk
Total Tweets Collected: 200
Geo Tweets: 0
Place Tweets: 0
High Risk Alerts: 0
Geocoded Tweets: (feature coming soon!)
9/5/2013
16:10:13
1 1 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
17:27:39
8 9 7 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/5/2013
19:14:28
5 8 6 Yes Yes
9/5/2013
20:12:06
2 8 8 Yes Undecided
9/5/2013
20:12:32
3 8 7 Yes Undecided

79

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/5/2013
21:03:57
3 10 7 Yes Undecided
9/5/2013
23:03:47
10 1 1 No No As someone who's a frequent user of location based services
(Foursqare) and one who enables geo location on my phone (and
various apps!) - you got me SOOOOO wrong!!
maybe if you increase your Twitter API limits (beyond 200 tweets),
you'll have a much better outcome.

Cheers
@razchorev
9/6/2013
2:33:19
3 7 3 Yes Undecided
9/6/2013
3:48:49
7 9 1 Yes Yes
9/6/2013
7:51:44
7 8 8 Yes Yes
9/6/2013
9:15:23
3 9 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/6/2013
9:26:19
2 9 10 Yes Yes I guess I forget to turn off my gps a lot. From this map it's pretty
obvious where my work and apartment is. Damnit... You should put a
direct link to deleting the tweets.
9/6/2013
9:57:45
5 10 7 Yes Yes
9/6/2013
10:57:26
3 10 10 Yes Yes I'm not too worried about sharing my location, even after seeing my
house lite up like a Christmas tree
9/6/2013
12:21:16
7 3 5 No Geosocial
Footprint did
not locate
areas of
concern for

80

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/6/2013
12:48:55
4 10 10 Yes Yes I didn't realize I was tweeting my location on most of these. To the
untrained eye these just look like places I like to hang out. Strangers
wouldn't know they're my house or work. My friends and family would,
but they already know this data anyways. I guess I'm confused at the
risk here?
9/6/2013
13:30:32
5 2 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/6/2013
13:47:09
5 10 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Didn't find my house, but the footprint is scary. I don't venture out of
that boundary often.
9/6/2013
17:41:15
8 8 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/6/2013
21:19:39
4 9 9 Yes Undecided
9/6/2013
21:51:40
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
While I was already aware of location privacy risks, I am interested in
this tool as a way of illustrating data privacy issues and think it has
potential for that purpose.
9/7/2013
1:49:13
2 8 10 Yes Yes Mostly was foursquare check-ins.

81

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/7/2013
10:21:17
5 10 10 Yes Yes Just made my tweets private after seeing this. I was high risk with 148
out of 199 and you knew where I worked and lived.
9/7/2013
16:52:04
10 10 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/7/2013
19:28:25
8 9 8 Yes No
9/7/2013
20:04:49
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
It showed nothing for my own account which I'm glad about as I do
like to keep my privacy :)
But I do volunteer work for a page and it showed a pretty precise
location for it etc
Interesting stuff...keep up the good work :)
9/7/2013
20:55:34
5 10 10 Yes Yes I share my location openly but I don't like seeing my house on here.
9/7/2013
23:20:02
1 1 1 No Geosocial
Footprint did
not
I do not think the website worked. I looked up a few different accounts
and none had any data
9/7/2013
23:50:09
7 9 5 Yes Yes As a quantitative geographer I'd like to see more people providing
anonymous coordinate information. There is a lack of information on
different types of movements and the processes that underlying
movement patterns.
9/8/2013
1:01:26
3 8 8 Yes Undecided I think a KML would be more user friendly as a download. What do I
do with a csv?
9/8/2013
1:43:59
3 9 9 Yes Undecided
9/8/2013
15:14:46
10 1 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
Within the first 45 mins of the Fed's becoming aware of a Yes
Men/San Diego Museum of Art hoax, they proved that they have
instant access to Twitter user accounts - by shutting down the
hoaxster's account - without warning or notification from Twitter.

82

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/8/2013
16:13:19
10 10 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/8/2013
19:15:59
3 9 8 Yes Undecided Can I put myself on a do not search list for your service? I don't want
other seeing my footprint.
9/9/2013
1:33:42
1 1 5 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
almost everyone i tried doesnt share GPS data so i actually havent
seen a result. everyone cant be found. Q3 is therefore a middle of the
road answer because it's not applicable.
9/9/2013
3:34:52
1 2 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/9/2013
10:32:19
3 8 8 Yes Undecided
9/9/2013
10:50:39
6 6 6 Yes Yes Nice service!
9/9/2013
10:55:49
4 4 8 Yes Yes
9/9/2013
12:56:33
8 7 5 Yes Undecided
9/9/2013
13:21:50
3 2 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Really great idea! Thanks! Though this time nothing concerning was
found for my account, I´m really looking forward to some of the new
features. I think being aware of how many informations someone can
get out of your tweets, even though you think you might be a "private"
person on Twitter, might help people understand how important
privacy and always being careful with what you share is!

83

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/9/2013
13:50:32
4 2 4 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/9/2013
17:10:22
2 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/9/2013
15:52:48
3 6 5 Yes Yes
9/9/2013
17:11:35
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for

9/9/2013
18:53:57
2 2 2 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Not sure if this is possible, but provide examples of accounts that are
crazy. Took me a while to find an account that actually did something
on the map. Was beginning to wonder if it worked.
9/9/2013
22:14:14
6 2 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
you can request the footprint/privacy risk of any twitter account, even
you are not the owner of the account - is this intended?

kind regards
9/10/2013
2:03:53
5 2 1 Undecided Geosocial
Footprint did
not locate
Interesting, but not very far developed yet

84

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/10/2013
4:40:58
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/10/2013
6:15:34
2 2 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I noticed one mistake: Despite me saying in my profile "glad to be in
Germany" (virtually only location I share), your algorithm apparently
didn't notice that but showed the map of the US! No big deal, but
should be improved in future versions, imho. Otherwise great idea,
thx for the hard work you invested!
9/10/2013
6:43:03
10 10 1 Yes No I consider it important, useful, and good etiquette to share my location
in my tweets.
9/10/2013
6:51:45
5 3 1 Yes No
9/10/2013
7:08:59
7 7 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I guess you just used the "coordinates" and "place" keys in your
analysis. Therefore showing me a big green "NO RISK" is somehow
misleading. It might be a good idea to let the users know more clearly,
that much more geoinformation can be extracted by semantic and
context analysis of their tweets.
Looking forward to use your app again, when further analysis is
available.
9/10/2013
7:34:15
2 2 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/10/2013
8:18:07
1 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for

85

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/10/2013
10:39:04
2 2 5 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
As a Europe based user, I had to zoom out from North America to
use. Might be better to start less geographically specific and zoom in!
9/10/2013
10:59:49
4 3 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
I was delighted to find that my social media paranoia has paid off and
I am (so far) not locatable via Twitter.
9/10/2013
12:23:50
6 6 6 Yes Yes
9/10/2013
16:22:13
3 9 8 Yes Undecided
9/10/2013
16:49:07
3 9 8 Yes No The areas of concern seem to be triggered by text. It would make
more sense to go off time and clustering. If you look at where I post at
night, you can see my house, yet it guesses my house is near my
office because I said I was on my way home?
9/10/2013
16:52:30
5 3 1 Undecided Geosocial
Footprint did
not
Regarding Q "How would you rate the personalized suggestions you
received on reducing your "
9/10/2013
19:38:04
4 3 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account
I thought the app would be able to detect my location based on some
other factor, aside from my already allowing location sharing on my
account. What this app actually helped reinforce was that I'm satisfied
with the amount of location sharing I do (which isn't much, about 10%
of tweets and when I do it's intentional and for promo or business). It's
plain dumb to broadcast personal stuff like home or where I'm eating
at the moment I'm eating there. When I broadcast location it's usually
when I'm about to leave. Safety first.
9/10/2013
23:39:02
3 3 10 Yes Geosocial
Footprint did
not locate

86

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/11/2013
0:02:08
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
This is more of interest (to me) as an open-source intelligence
gathering tool.
9/11/2013
10:54:53
6 1 10 Yes Yes
9/11/2013
12:02:35
8 4 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/11/2013
15:07:58
7 7 8 Yes Yes
9/11/2013
20:09:20
3 9 8 Yes Undecided
9/12/2013
4:41:55
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/12/2013
4:42:21
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/12/2013
8:30:22
3 3 3 Yes Geosocial
Footprint did
not locate
areas of
I don't use geoinformation myself but i;\'ve looked at a few friends and
i'm amazed about what i can fnd. Some of them were lying about
there whereabouts in tweets. So the tool has a lot of effect. I'll never
never use geoinformation in twitter myself.

87

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/12/2013
9:16:33
10 10 1 Undecided No Sorry man, but it doesn´t works.
9/12/2013
17:48:26
7 4 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
0:55:30
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Extremely fascinating and useful piece of programming: This could be
used by individuals and/or organizations to check the Geossocial
Footprint of members or friends and give them helpful warnings.
9/13/2013
2:09:07
2 8 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I'm waiting for the tweet context analysis, geocoded tweets, hoped
GF would be able to see through for used IP addresses for example
to locate tweeps.
Tried a few other tweeps like @4positiviteit who has Nijmegen in the
bio and no report at all for that and @jacquessmits who got :ow Risk
4 location ratings - twitters complete addresses zipcode and numbers,
@BCoachOpleiding who twitters cities, @cpap73 city in bio,
@Beddoloog complete address in bio, @chgroenusa city in bio, none
of them reported. @w8eens who tweets with GPS location was
reported adequately high risk.
So it seems to work fine!
9/13/2013
7:54:43
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Only No Help as I had no items of concern to advise me on
9/13/2013
11:32:29
1 1 3 Yes Geosocial
Footprint did
not locate
I have mentioned some names of towns and villages I visited and
near my home.

88

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/13/2013
11:40:09
1 1 8 Yes Yes I just gave feedback that the single Geotag I've made was not shown.

It worked the second time.
9/13/2013
11:51:28
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
11:55:07
1 1 1 No Yes
9/13/2013
12:09:24
4 4 7 Yes No Well, I am well aware when I share location information, so the tool
did not change my location privacy assessment. As I do not tweet
often, the "above average" message is no concern to me. When
zooming in, I noticed that the four located areas of concern in London
were neither the place where I live, nor the place where I work. I think
it's a great tool, especially perhaps for people who do not notice when
they share their locations!
9/13/2013
12:33:12
10 10 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
12:41:46
1 1 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
12:44:41
10 10 1 No No
9/13/2013
12:48:24
2 2 1 Yes Yes

89

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/13/2013
13:48:48
3 3 4 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Since I only had 2 locations, I had no tips..
Nice app!
9/13/2013
13:20:39
7 7 1 Undecided Undecided i wasnt able to get any information on the twitter usernames I entered
9/13/2013
13:48:57
3 7 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
14:09:28
5 5 2 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
15:05:51
1 1 4 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I use location terms like city or venue names in my tweets, but they
wasn't detected by geosocial footprint. But with this website I'm now
more aware about privacy issues even though I thought Im careful.
Thx
9/13/2013
15:41:53
5 10 6 Undecided Yes Good approach, but missing hashtag search (twitter is famous for
this).

Greetz from Austria.
9/13/2013
16:04:32
2 2 1 Yes Geosocial
Footprint did
not locate
areas of

90

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/13/2013
17:40:31
1 1 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
18:01:22
3 2 2 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
19:28:19
8 4 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/13/2013
21:51:45
6 1 5 Yes Undecided
9/14/2013
1:37:47
3 4 5 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
tried 6 accounts, non got results
9/14/2013
2:16:13
3 2 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/14/2013
4:46:09
1 1 10 Yes Geosocial
Footprint did
Good work! Best regards from Germany

91

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/14/2013
5:03:24
5 5 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/14/2013
5:19:47
6 4 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/14/2013
5:23:37
8 8 1 No No unfortunately it did not work
9/14/2013
11:45:23
1 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/14/2013
12:31:45
2 2 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/14/2013
14:39:11
5 1 1 Undecided Geosocial
Footprint did
not

9/15/2013
4:59:54
2 6 10 Yes Yes I was surprised by the results... I will get my family to check it out too.
9/15/2013
5:56:16
3 2 1 Yes Geosocial
Footprint did
not locate

92

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/15/2013
6:23:06
3 1 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great product!
9/15/2013
8:48:36
6 10 8 Yes Undecided
9/15/2013
9:13:12
6 6 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/15/2013
9:32:04
6 4 9 Yes Yes
9/15/2013
9:58:45
10 10 5 Undecided Undecided
9/15/2013
10:00:59
3 8 8 Yes Yes looks like it only uses english. plus add others -Germany
9/15/2013
10:28:56
6 4 9 Yes Yes
9/15/2013
10:42:33
5 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/15/2013
12:39:11
2 5 5 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

93

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/15/2013
13:55:58
2 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/16/2013
3:09:11
10 5 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/16/2013
8:06:42
4 4 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account
I had no results
9/16/2013
14:44:05
5 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/16/2013
17:32:28
8 3 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Didn't have any personalised suggestions as it said account was ok.

However, I'd spent last week at a conference, using a # tag a lot. It
might be useful in future iterations to see if it's possible to link tags &
locations, for, while not as definite as a location set by the phone
etc., lots of tags from a particular location could indicate possible
locations, tho guess you'd need access to multiple Tweeters.

Hope that makes sense - writing on tablet :-)

94

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/18/2013
11:17:12
1 1 7 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Maybe try to include Facebook check ins etc? Although I would be
inclined not to use it if j had to log into Facebook, so actually no, don't
do this
9/18/2013
12:28:30
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Interesting.
9/19/2013
3:17:01
7 10 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
tryed with @u14183 and nothing was happen.
9/19/2013
6:57:37
5 5 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/19/2013
7:56:20
10 6 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
A fascinating tool
9/19/2013
16:16:25
3 3 5 Undecided Geosocial
Footprint did
not locate
areas of

95

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/20/2013
6:40:04
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/20/2013
8:23:38
3 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Potentially useful tool - without some feedback on sharing risk (just
because I had no sharing information to provide feedback on) it's hard
to assess the value.
9/23/2013
15:42:46
6 8 8 Yes Yes
9/23/2013
22:37:14
5 5 8 Yes No Very good tool
9/24/2013
2:30:41
4 4 10 Yes Yes Heatmap is good GUI
9/24/2013
17:01:29
5 5 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/24/2013
17:01:32
5 5 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

9/26/2013
10:40:13
8 10 8 Yes Yes
9/26/2013
14:47:19
1 1 10 Undecided No I wasn't able to locate anybody even open accounts

96

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
9/27/2013
14:07:37
5 4 1 Undecided Undecided My location is clear from some of my geographically-referenced
tweets whether or not the tweet is geo-tagged.
9/27/2013
14:16:20
10 10 1 No Yes of the two major geolocations shown, one was wildly inaccurate.
9/28/2013
17:39:42
8 8 1 No No
9/30/2013
15:53:05
6 5 10 Yes Yes bu harika bir şey
10/2/2013
3:25:40
2 2 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
I'm very selective about turning location on. The only tweets were
purposely geolocated for the USC Spatial Science Catalina Field trip
to show places of interest. So while they show I was spending time at
locations on Catalina Island, they show no patterns of tweeting or
reveal workplace or home. That being said, I was pleased to see that
nothing else showed up on the map.
10/2/2013
11:27:52
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

10/4/2013
11:10:09
6 6 7 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

10/6/2013
8:00:13
3 8 8 Yes Undecided I saw areas of concern but not sure how to check if they were
accurate because I don't know the GPS location of my house. Maybe
give a street address?
10/6/2013
8:01:35
4 9 9 Yes Yes Nice article. This is kind of scary. While I personally don't have any
locations on the map, my kids each have our home and their local
hangouts visible. Thanks for helping me be a better parent.
10/7/2013
3:37:36
3 2 5 Undecided Undecided

97

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
10/7/2013
4:05:24
8 8 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Develop its feature further to also check for hashtags
10/7/2013
8:59:48
3 9 10 Yes Undecided I didn't realize I was sharing my location so often. Does it collect
location automatically on some tweets?
10/7/2013
9:37:16
6 6 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
No info or tweets showed on map?
Windows phone incompatible?
Interested in trying this out.
Cindy Forbes @getmeoutnews
10/7/2013
11:28:06
4 10 8 Yes Yes Who knew that saying Fuck School would be enough for someone to
figure out where I go to school.
10/7/2013
15:57:20
5 4 7 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

10/7/2013
16:50:45
4 9 7 Yes Undecided I'm not sure what to think. It showed a few areas of concern. Most of
them were wrong but one was marked as my home, but instead it was
my university. I assume this means I'm spending too much time there
or something?
10/7/2013
19:11:01
1 1 8 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

10/8/2013
9:44:34
4 9 9 Yes Geosocial
Footprint did

98

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
10/8/2013
9:45:07
4 8 9 Yes Undecided
10/8/2013
10:04:49
4 8 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

10/8/2013
12:52:21
4 8 8 Yes Undecided
10/8/2013
16:00:03
4 8 8 Yes Undecided
10/9/2013
13:13:19
3 8 9 Yes Undecided Some of the areas of concern were remotely accurate but most were
wildly incorrect. I don't live on the freeway.
10/9/2013
20:29:04
4 9 8 Yes No
10/10/2013
8:03:19
5 9 9 Yes Undecided
10/10/2013
10:36:41
1 5 5 Yes Yes
10/10/2013
15:20:31
2 2 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Thanks!
10/11/2013
13:53:25
6 6 8 Yes No I appreciate the educational component, but you could easily
demonstrate this using a hypothetical person. Your application
seems unethical from an educational standpoint because it provides
an easily available tool to gain public, yet hard to find, information to a
broad audience.
10/11/2013
15:36:03
5 6 7 Yes Yes

99

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
10/11/2013
16:34:47
4 9 8 Yes Undecided
10/13/2013
20:55:17
3 8 8 Yes Yes 1 of the 10 or 11 areas of concern were accurate. I am not sure if that
is a good thing or a bad thing.
10/14/2013
15:40:44
5 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

10/15/2013
11:54:31
5 9 8 Yes Undecided
10/15/2013
16:18:10
5 10 8 Yes No
10/17/2013
10:06:15
1 1 1 No Geosocial
Footprint did
not locate
areas of
concern for
my account

10/20/2013
9:07:28
10 10 10 Yes Yes son derece yararlı bir uygulama bence
10/22/2013
15:39:16
1 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

10/23/2013
21:53:38
4 9 4 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

100

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
10/24/2013
9:33:39
1 1 9 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

10/24/2013
9:55:14
7 7 6 Yes Yes
10/27/2013
12:53:47
10 10 5 Yes Undecided
10/27/2013
16:10:45
10 5 7 Yes Undecided I'm happy to provide my location in tweets... helps people in the same
area connect.

I can see however how it could be a privacy concern when you tweet
most of your tweets from home or office... there's crazies out there
who might just use that info..
10/27/2013
16:15:28
10 10 3 Yes Yes Mhmm,... thinking about it again. This is very very helpful is showing
the risk... but not helpful at all
10/28/2013
13:32:47
1 1 1 Yes Undecided
10/30/2013
4:28:32
5 2 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
would be good to develop features that alert users when location
information has been identified.
10/30/2013
19:28:30
1 1 1 Undecided Undecided
10/31/2013
2:46:28
9 9 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

101

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
11/6/2013
20:36:47
10 10 10 Yes Yes
11/7/2013
0:55:20
2 4 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Good idea, thanks for making this!
11/9/2013
7:01:48
4 2 4 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
too soon.
I was requested for this questionaire before I even had a chance to
play with the website!
11/9/2013
8:26:04
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

11/9/2013
16:50:38
6 9 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

11/11/2013
3:31:08
6 8 8 Yes Undecided
11/13/2013
16:12:47
1 1 1 Yes Undecided
11/14/2013
16:34:11
8 8 3 Yes Yes
11/14/2013
17:13:55
5 6 6 Yes Yes

102

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
11/15/2013
16:52:31
2 3 4 Yes Undecided
11/16/2013
7:44:30
7 7 7 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
looking forward to more options
11/21/2013
12:30:42
4 2 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

11/22/2013
6:36:33
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

11/24/2013
22:39:31
1 1 5 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

11/27/2013
5:01:49
2 7 8 Yes Yes
11/28/2013
9:30:55
2 2 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
I did tweet about you to teachers and mediacoaches.

103

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
11/29/2013
13:56:41
8 6 4 No Yes
12/2/2013
15:18:26
4 1 6 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

12/9/2013
9:50:11
1 1 10 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account

12/9/2013
16:03:16
1 1 1 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Great app, think that IOS users have a higher level of security by
default and android more open.
I rated the personalised suggestions as no help as I already had my
account locked down
12/11/2013
8:22:53
8 6 3 Yes Yes
12/13/2013
10:12:27
5 5 5 Yes Yes
12/14/2013
3:03:17
1 1 5 Yes Geosocial
Footprint did
not locate
areas of
concern for
my account
Cool app
12/17/2013
6:14:52
3 2 8 Yes Geosocial
Footprint did
not locate
areas of
useful tool. Thankyou

104

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
12/22/2013
7:42:56
4 4 1 No Undecided For 5 twitteraccounts i found no risks at all
12/28/2013
8:06:33
8 10 9 Yes Yes
1/6/2014
11:30:50
8 9 7 Yes Yes
1/7/2014
19:09:17
4 1 5 Yes Yes
1/12/2014
21:59:57
8 8 7 Yes Yes
1/9/2014
16:15:57
10 10 8 Yes Yes
1/13/2014
5:46:53
10 3 10 Yes Yes
1/14/2014
6:10:59
5 10 7 Yes Yes
1/14/2014
14:02:55
1 1 6 Yes Yes
1/16/2014
0:07:08
5 5 7 Yes Yes
1/17/2014
12:31:40
1 1 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

1/20/2014
2:23:26
9 10 10 Yes Yes
1/20/2014
12:47:01
5 5 5 Yes Undecided
1/22/2014
14:05:51
3 6 8 Yes Yes Great tool! I am a geographer and will be using this in a course where
students map their movements over the course of the semester using
anonymous twitter accounts to explore how they interact with the

105

Timestamp Prior to
using
Geosocial
Footprint,
how would
you have
rated your
location
privacy on
Twitter.com?
After using
Geosocial
Footprint,
how do you
rate your
location
privacy on
Twitter.com?
How would
you rate the
personalized
suggestions
you received
on reducing
your over
sharing
risk?
Did you find
Geosocial
Footprint
educational?
If Geosocial
Footprint
located areas
of concern,
were any of
them
accurate?
General comments?
1/22/2014
22:58:37
3 1 1 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account
I don't think it is entirely accurate. I tweeted about being at the Detroit
Auto Show, as well as being at home or going to a hockey game, all
things which could tell people where I am.
1/23/2014
14:40:22
7 3 8 Undecided Geosocial
Footprint did
not locate
areas of
concern for
my account

1/23/2014
15:57:48
6 6 5 Yes Undecided
1/24/2014
11:30:27
1 10 2 No Undecided
1/25/2014
21:51:53
4 2 7 Undecided Yes
1/29/2014
11:46:05
3 2 3 Yes Yes

Asset Metadata

Creator Weidemann, Christopher Donald (author)

Core Title GeosocialFootprint (2103): social media location privacy Web map

Contributor Electronically uploaded by the author (provenance)

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 04/29/2014

Defense Date 03/13/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag geographic information systems,geosocial,geospatial,GIS,Google API,heatmap,Javascript,OAI-PMH Harvest,Twitter

Format application/pdf (imt)

Language English

Advisor Swift, Jennifer N. (committee chair), Knoblock, Craig (committee member), Pultar, Edward (committee member)

Creator Email cdweidem@usc.edu,chrisweidemann@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-406404

Unique identifier UC11296688

Identifier etd-WeidemannC-2456.pdf (filename),usctheses-c3-406404 (legacy record id)

Legacy Identifier etd-WeidemannC-2456.pdf

Dmrecord 406404

Document Type Thesis

Format application/pdf (imt)

Rights Weidemann, Christopher Donald

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract Spatial thinking is an abstract term and process in regards to what most of the general population understand. Many people are not well versed in geospatial terminology, options of use, and the location intelligence they unconsciously disclose when using social media outlets. This thesis integrates a unique technical web application with GIScience intended to illuminate the subsequent effect location‐based data can have on one’s personal privacy, security, and web‐presence. An innovative new web mapping application was built for general public consumption that aggregates location data from Twitter, harvests ambient location information, analyzes the captured data to provide personal location intelligence, and visualizes possible areas of interest. In addition, the research examines the results of an online voluntary survey collected from the users of the application. Finally, this thesis discusses how these same techniques can be applied to other social media outlets along with potential opportunities to educate and inform the general public more about their social media location privacy.