University of Southern California Dissertations and Theses

Finding your best-fit neighborhood: a Web GIS application for online residential property searches for Anchorage, Alaska

(USC Thesis Other)

Finding your best-fit neighborhood: a Web GIS application for online residential property searches for Anchorage, Alaska

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

FINDING YOUR BEST-FIT NEIGHBORHOOD:
A WEB GIS APPLICATION FOR ONLINE RESIDENTIAL PROPERTY SEARCHES
FOR ANCHORAGE, ALASKA

by

Jennifer Nicole Dowling

__________________________________________________________________

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)

December 2014

Copyright 2014 Jennifer Nicole Dowling

ii

DEDICATION
To my ever-patient husband, Paul, who encouraged me to realize my potential.

iii

ACKNOWLEDGMENTS
Profound and numerous thanks go to everyone who supported me with scholastic and future
employment advice, computer programming solutions, and who lent themselves as a sounding
board during the more challenging aspect of this project. To Dr. Jennifer Swift, who provided
constant encouragement and ready and invaluable assistance. To Dr. Su Jin Lee, who gave sharp
insight coupled with sly wit early in my program. To Dr.Yao-Yi Chiang, who showed simple
solutions to complex database design so useful to my project. To Mr. Murphy, for his laws that
seemed to influence my project greatly. I salute you all.

iv

TABLE OF CONTENTS
DEDICATION ii
ACKNOWLEDGMENTS iii
LIST OF TABLES vi
LIST OF FIGURES vii
LIST OF ABBREVIATIONS x
ABSTRACT xi
CHAPTER 1: INTRODUCTION 1
1.1 Motivation 1
1.2 Research Questions 2
1.3 Hypothesis 3
CHAPTER 2: RELATED WORK 5
2.1 Web-Based Real Estate Sites 5
2.2 Online Web Real Estate Consumer 13
2.3 Caveats to Neighborhood Search Models 17
CHAPTER 3: METHODS AND DATA SOURCES 20
3.1 Study Area 20
3.2 Web Application Development 21
3.3 Source Data 28
3.3.1 Source Data for Hazards Theme 29
3.3.2 Source Data for Education and Recreation Theme 42
3.3.3 Source Data for Political Theme 45
3.3.4 Source Data for Transportation and Zoning Theme 47

v

3.3.5 Source Data for Housing Theme 49
3.4 Data Configuration in the Find Your Anchorage Neighborhood Application 54
3.5 The Web-Based Design Application Development 58
3.5.1 Web Application Planning 59
3.5.2 Web Application Development Efforts 60
3.6 Final Web Map Programming 61
CHAPTER 4: RESULTS 67
4.1 A Tour of the Find Your Anchorage Neighborhood Application 67
4.2 Challenges Encountered 71
4.2.1 Esri ArcGIS and Microsoft SQL Server 71
4.2.2 HTML and JavaScript 73
CHAPTER 5: CONCLUSIONS 77
5.1 The Future of Real Estate Search Websites 77
5.2 The Ever Expanding Future of Web GIS Map Programming 78
5.3 Future Work on the Find Your Anchorage Neighborhood Application 79
REFERENCES 82

vi

LIST OF TABLES
Table 1 Source Material Information (Municipality of Anchorage 2013) 22
Table 2 2012 "Calls For Service" Crimes By Type in the Anchorage Bowl (Morberg 2013) 23
Table 3 All Layers That May Be Viewed or Incorporated in a Search 24
Table 4 "Selection Package" Search Feature Services 57

vii

LIST OF FIGURES
Figure 1 Results of the Interactive Filter from Homes.com 5
Figure 2 Result of Homes.com Website Search Filter 6
Figure 3 Realtor.com Amenties Legend Near a Property in Lexington, Kentucky 8
Figure 4 Available Schools Toggle Using Moriches, NY Data from Zillow.com 9
Figure 5 Selection of Chart Data on Redfin.com for Miami, Florida 9
Figure 6 Trulia.com Selected Local Data for Bellingham, Washington 10
Figure 7 Trulia.com Display of Commute Times in Phoenix, Arizona 10
Figure 8 Local Amenities Near a Property in Kansas City, Missouri From Trulia.com 11
Figure 9 RaidsOnline.com Display of Anchorage, Alaska 12
Figure 10 Top Real Estate Vertical Matrix 14
Figure 11 Extract of Neighborhood Description of Ironwood, Michigan 15
Figure 12 Graphic Demographic Representation of a Neighborhood in Ironwood, Michigan 15
Figure 13 Filtering Options for Manhattan, Kansas 16
Figure 14 Property Information from Homefacts.com 16
Figure 15 Realtor Links for a Zillow Listing 18
Figure 16 The Anchorage Bowl and Vicinity (Municipality of Anchorage 2013) 21
Figure 17 Flowchart Showing the Find Your Anchorage Neighborhood Creation Process 26
Figure 18 Example of Crime Types by Date and Address in ArcMap. Note the Sexual Assault
Addresses are Relisted as APD Headquarters. 30

viii

Figure 19 Crime Points Plotted at the 100-Block Level 31
Figure 20 Classifying and Coloring the Heat Map 33
Figure 21 Anchorage Bowl Violent and Nonviolent Crime in 2012 (Morberg 2013) 34
Figure 22 Anchorage Crime Heatmap 35
Figure 23 Home or Employment Locations of Registered Sex Offenders (Alaska Department of
Public Safety 2013) 37
Figure 24 Seismic Layers in the Anchorage Bowl (Municipality of Anchorage 2013) 39
Figure 25 FEMA 100-Year Flood Plan (Federal Emergency Management Agency 2014) 41
Figure 26 All Anchorage School District Schools and School District Boundaries (GreatSchools
2013) 43
Figure 27 MOA Trails and Parks (Municipality of Anchorage 2013) 44
Figure 28 Political Theme (Municipality of Anchorage 2013) 46
Figure 29 Transportation and Zoning Theme (adapted from (Municipality of Anchorage 2013)48
Figure 30 MOA Zoning Codes By District Type from (Municipality of Anchorage 2013) 49
Figure 31 Housing Theme (Municipality of Anchorage 2013) 50
Figure 32 Excerpt from the Available Housing Attribute Table 51
Figure 33 A Size Comparison of Neighborhood Versus Subdivision Layers 53
Figure 34 Data Preparation Tasks Performed on Trails, Parks, and Seismic Layers 54
Figure 35 Layers Created From Zoning Layer 56
Figure 36 Seismic Layer URL 58

ix

Figure 37 High Level Flowchart of Web GIS Application Creation Steps 61
Figure 38 Public Information Template 62
Figure 39 ArcGIS.com Map With Webmap ID 64
Figure 40 Great Outdoors Selection with Descriptive Popup Window 65
Figure 41 The Swipe Display Option in the Web Application 66
Figure 42 The Three Tabs in the Application Expanded 68
Figure 43 Partial Metadata for the Find Your Anchorage Neighborhood Web GIS Application 69
Figure 44 One of Several 1600 Pennsylvania Avenues 70
Figure 45 Social Media Options 70
Figure 46 Insufficient Space Allocation Error in ArcMap 73
Figure 47 Drop Down Menu Option in Trulia.com 75
Figure 48 Finding the Layer Text ID Using Firebug 76
Figure 49 Crime Mapping Display From Trulia.com 80

x

LIST OF ABBREVIATIONS

AMD Asynchronous Module Definition
API Application Program Interface
APD Anchorage Police Department
ASD Anchorage School District
CFS Calls for Service
DPS Alaska Department of Public Safety
ESRI Environmental Systems Research Institute, Inc.
FEMA Federal Emergency Management Agency
GIS Geographic Information System
GIST Geographic Information Science and Technology
HTML5 HyperText Markup Language, revision 5
IIS Internet Information Services
KML Keyhole Markup Language
KMZ Zipped Keyhole Markup Language
MLS Multiple Listing Service
MOA Municipality of Anchorage
MOA GIS MOA, Information Technology Department - GIS Services
NAD83 North American Datum of 1983
REST Representational State Transfer
USC University of Southern California

xi

ABSTRACT
Online geospatial data are evident in many websites, covering a variety of interests such as route
planning, incident locations, and outdoor recreation searches. One type of geospatial website is
the online real estate search. Many realty websites allow prospective residential property buyers
to sort listed properties interactively based on desired elements. These elements typically address
features wanted within a home, such as the dwelling's size, number of bedrooms and bathrooms,
whether a garage or swimming pool is included, and other furnishings. Equally important in
considering the ideal home is to find the ideal location. Length of the commute time, crime
frequency, proximity to cultural and retail options, and the location of desired schools can
provide for an overall "neighborliness" that is vital to ensure a comfortable life in the new home
at the new location. While some websites are beginning to address this concern by including
small overlays within a property's webpage, none overtly considers that the home buying process
may not start with the selection of home features, but by first determining the "best-fit"
neighborhood. The web application created for this thesis is unique in its premise to first
introduce potential homebuyers to neighborhoods. Prospective homebuyers may select from
several neighborhood factors to find locations that satisfy their search parameters. An overlay of
available properties is then displayed for the web application user to show what offerings are
available in those resulting areas.

CHAPTER 1: INTRODUCTION
This research project is designed to create a web application to allow prospective residential
homebuyers to begin an online home search by first considering elements that make for a
desirable neighborhood. The main goal of this project was first to understand the dynamic nature
of geospatial website programming as it can be used to integrate neighborhood-based search
features into existing residential real estate websites. The resulting web application is intended to
provide a distinctive search for both the prospective buyer and his or her realtor by incorporating
neighborhood-level geospatial data for the pilot study area of Anchorage, Alaska, in a unique
approach that may be replicated in other municipalities.
1.1 Motivation
Home purchases are typically the largest expense one can make (Hoak 2010), and its
selection is often an emotional one. Finding a home with the correct number of bedrooms and
bathrooms is easy to do when researching real estate listings online, but it is unusual to
incorporate an exploration of comfortable environments into the same general search. A good
home in a less-than-ideal neighborhood can be disappointing.
One popular use for online geospatial data involves residential real estate, where
prospective homebuyers can direct searches of available properties from a number of websites
even before engaging a realtor (Trefethen 2013). According to the National Association of
Realtors, 92 percent of homebuyers searched internet resources when searching for property and
for 42 percent of buyers, examining online real estate websites was their first step (National
Association of Realtors 2013). For the purposes of this thesis research, a property is defined as a
residential dwelling, such as a single-family house or townhome that is available for purchase.

2

Real estate agents that present detailed information on available homes and land can realize
increased interest in their properties and thus increased sales (Benjamin et al. 2005). This thesis
web geographic information system (GIS) application was designed to assist two types of home
shoppers: the recent transplant that may be unfamiliar with the town and is not acquainted with
the various neighborhoods, and the long-time resident who wants to relocate but may not know
the intricate social and environmental differences among the various sections of town. While an
effective realtor can provide knowledgeable advice about the area, a comprehensive real estate
website can also allow the buyer to research neighborhoods as simply as many websites currently
allow general property research. Currently, most real estate websites pay little attention to these
elements in the surrounding neighborhood. Are the schools good? Is the crime rate low? How far
away from major roads, airports, or commercial properties is desired? Are there any natural
hazards that would affect safety or make obtaining property insurance difficult? These additional
elements must be considered at some point when looking for a home, and ideally, they should be
considered early in the home search process.
1.2 Research Questions
This thesis answers several research questions. Is it possible to create a web application
that contains the ability for users to filter neighborhoods by select criteria? Will such a web GIS
application be easy to navigate and to understand? Herein, for the purposes of this thesis, a
neighborhood is defined as a geographical subset of a municipal area that has characteristics
distinct from surrounding areas. These characteristics may be as simple as major road boundaries
or as complex as bounded social or ethnic associations. In this thesis, neighborhoods are smaller
units that satisfy homebuyers' proximity desire for recreation, safety, or commerce.

3

According to real estate website site traffic numbers, current online real estate searches
are a popular and practical means to shop for such large expenses. In February 2014, 316 million
visitors went to an online real estate website from their desktop computer (Hagey 2014). Those
numbers do not necessarily indicate a homebuyer or a unique visitor (they could be realtors,
other home sellers, or real estate website developers), but they do suggest a viable business
model. Online research is not confined to the desktop computer; the two largest websites,
Zillow.com and Trulia.com, estimate that mobile device users account for almost half their
website traffic (Hagey 2014).
1.3 Hypothesis
A unique web GIS application may be created that emphasizes neighborhood importance
before home features. The application should be easy to navigate and intuitive, offer specialized
neighborhood-related search functionality, and be simple to maintain. Finally, such an
application can be constructed using select local knowledge rather than a more generic national-
based assemblage of data, to fully inform the visitor at a local scale.
The web GIS application developed for this thesis project is unique in online home real
estate searches compared to current online websites. This web GIS application offers the
following local information and functionality, most of which are not currently found in exiting
real estate websites:
 Trail routes with detailed use information
 Zoom and search within neighborhoods
 Political districts and current office holders
 Websites of all public schools

4

 Search by seismic layers
 Four prepackaged searches to assist in pinpointing optimal home search locations
 Ability to search by local address
 Share the resulting map via social media

5

CHAPTER 2: RELATED WORK
A web-based approach to offering real estate is neither unique nor rare, as shown by the number
of real estate websites available. National real estate companies (e.g., Remax.com,
Century21.com), local realtors, and businesses that display real estate offerings from multiple
agencies or sellers (e.g. Zillow.com, Redfin.com, and Realtor.com) create these particular
websites. All provide interactive maps that show available properties that the website visitor may
filter based on their specific requests.
2.1 Web-Based Real Estate Sites
Figure 1 shows a filter on the website Homes.com. In this example, the user selects a
3-bedroom, 2-bathroom single-family home in Key West, Florida, priced between $600,000 and
$800,000 and built sometime between 1900 and 1970. From those results, the visitor can click on
any interesting properties for more information, photographs, and listing agent contact
information, as seen in Figure 2.

Figure 1 Results of the Interactive Filter from Homes.com

6

Figure 2 Result of Homes.com Website Search Filter

7

Real estate web programming is often patented to secure intellectual ownership. Several
processes in early real estate website programming were submitted to the US Patent Office to
secure intellectual property rights for their authors (Choziak 2003); (Eraker, et al. 2005);
(Florance, et al. 2007); (Hancock 2003); (Hoffman, Anzalone, and Cormack, 2003). The ideas
presented in the patent applications suggest effective yet generic programming tactics. Eraker, et
al., (2005) as one of the first developers submitted a patent application for the residential online
real estate company Redfin.com that outlines a web-based program to include aerial image
overlays and the ability to search various databases, including tax rolls, multiple listing service
(MLS) listings (Multiple Listing Service 2012), and local government GIS datasets (Redfin
2012). These processes collect the available properties to create their databases and in turn build
the website. Such early real estate websites served as examples for subsequent sites that use
geospatial techniques to locate housing options.
Several national-level real estate websites including Realtor.com, Redfin.com, and
Zillow.com also provide local demographic or financial data either as a general overview of a
town or for user-selected properties. Realtor.com also displays limited local information when
the user selects a property from the map (Figure 3). Zillow allows users to choose school
boundaries as a neighborhood filter (Figure 4), and helpfully links school performance data from
GreatSchools.com (GreatSchools 2013). Many other websites do not show any local information,
at least not in map form or in otherwise readily searchable formats. Instead, Redfin uses a
number of charts to describe the demographics, climate, and housing costs by zip code (Redfin
2012). Although this is useful information, it is not actionable when users want to quickly sort
properties in a map (Figure 5).

8

One of the more innovative companies is Trulia.com. Their website employs a broad
selection of local information (Figure 6). Users can view overlays of local amenities, school
boundaries, floodplains, and earthquake zones, and via a simple interval map see travel times
from a selected origin (Figure 7). Trulia has data for most of the United States; for locations
where they do not have comprehensive coverage, users can at least view school boundary
polygons and their associated Greatschool.com ratings. Similar to the other websites, Trulia
displays these data layers from the city overview page, or in limited form from a selected
property's page rather than as an introduction to a neighborhood search (Figure 8) (Trulia 2012).

Figure 3 Realtor.com Amenties Legend Near a Property in Lexington, Kentucky

9

Figure 4 Available Schools Toggle Using Moriches, NY Data from Zillow.com

Figure 5 Selection of Chart Data on Redfin.com for Miami, Florida

10

Figure 6 Trulia.com Selected Local Data for Bellingham, Washington

Figure 7 Trulia.com Display of Commute Times in Phoenix, Arizona

11

Figure 8 Local Amenities Near a Property in Kansas City, Missouri From Trulia.com

Crime information is also often considered in a home search and is addressed not only by
the three big realty websites, but also by dedicated crime information websites such as
Raidsonline.com (BAIR Analytics 2011). This site receives crime data from participating law
enforcement agencies throughout the North America via automated feeds and populates the map
with both point data and heat maps (Figure 9). This interface also allows users to choose crime
types and buffer them around a location.

12

Figure 9 RaidsOnline.com Display of Anchorage, Alaska

The information in these websites is very useful to identify features of concern in an area,
but they are not prominent (easily accessible) in the property-level (individual home) search, and
they do not answer all the questions a prospective buyer may have, particularly about local
political boundaries or other demographic data. Moreover, because they either reference a
specific property or give a general overview of an area, the functionality of these real estate
search applications as detailed above cannot help homebuyers sort a town into neighborhood

13

choices. Whereas the website application created for this thesis project is designed to first guide
the user to search and find ideal neighborhoods based on a variety of layer selections, such as
business or industry, major roads, schools and recreation. From within the resulting boundaries,
users can then investigate available properties.
2.2 Online Web Real Estate Consumer
According to a study by the National Association of Realtors, ninety percent of
homebuyers begin their search online (National Association of Realtors 2013). As more real
estate websites come online and offer varied options to attract web traffic, home shoppers have
responded. In March 2014, the top twenty online real estate websites showed 22.4 million unique
visitors using either a desktop or mobile application, as shown in Figure 10. Of those, the top
three most-visited websites include Realtor.com, Zillow, and Trulia. These websites claimed
over half that amount (National Association of Realtors 2014). These numbers have increased
steadily indicating that this method is viable for introducing properties to prospective buyers,
perhaps regardless of buyer age. It is interesting to note that Zillow plans to purchase Trulia in
2015, and intends to operate both websites as independent businesses. (Light 2014).
As these real estate website flourish, specialized websites that provide various levels of
in-depth details about neighborhoods, similar to this thesis, also appear. Such sites include
NeighborhoodScout.com and Homefacts.com, where local information is prominent over
property searches. NeighborhoodScout (NeighborhoodScout 2014) is a membership site that
offers over 150 search parameters, most culled from US Census datasets, while Homefacts
(Homefacts 2014) offer only neighborhood snapshots and limited filtering options. Both sites
will link to local realtors, but neither provides home searches as its primary function.

14

Figure 10 Top Real Estate Vertical Matrix
Source: (National Association of Realtors 2014)
NeighborhoodScout delivers the results of a selection as a series of responses in
paragraph form and as charts, as seen in Figure 11 and Figure 12. Homefacts lets users identify a
variety of elements and a small selection of available properties within a given area (Figure 13).
Once a property is selected, Homefacts will show a snapshot of "local area highlights" such as
schools, crime rating, or natural hazards (Figure 14). Both websites offer the user a good choice
of options to narrow a home search and explain their parameters. A cursory examination of
Homefacts shows elements that may concern a homebuyer, such as placing Manhattan in the
wrong county (indicating geocoding problems) and inflating the importance of a category (listing
"Tanks and Spills" implies all the containers were hazards). Websites such as these are valuable,
but because they are designed to apply to the entire US theses websites end up being generic in
their search functionality and resulting data descriptions, lowering the amount of detail on
provided on a given topic. The homebuyer can benefit from a more local approach, which is
attempted in this thesis project.

15

Figure 11 Extract of Neighborhood Description of Ironwood, Michigan
Source: NeighborhoodScout.com

Figure 12 Graphic Demographic Representation of a Neighborhood in Ironwood, Michigan
Source: NeighborhoodScout.com

16

Figure 13 Filtering Options for Manhattan, Kansas
Source: Homefacts.com

Figure 14 Property Information from Homefacts.com

17

2.3 Caveats to Neighborhood Search Models
The number of options a shopper can integrate into a neighborhood search can be a
double-edged sword in how the output data may be used by consumers. Searches that incorporate
schools, commute times, and distances to services are simple, but adding granular information
from census data such as ethnic makeup, income levels, and religious affiliation can be
problematic. Such searches can be innocuous - families who want to live surrounded by other
immigrants from the same country, or religious sects who commonly live in community settings
- or they can appear more exclusionary and xenophobic. Real estate agents tend to avoid
addressing such search requests for fear of conflicting with fair housing laws (Wiggin 2014).
"Hyper local" data, while appealing to the more tech-savvy real estate shopper, can devolve a
home search into a series of frustrating results that yield no successful neighborhoods, and
therefore no properties of interest. In fact, while NeighborhoodScout uses over 100 search
parameters culled from census data and local government websites, none can reliably inform a
shopper whether the couple next-door keeps a team of howling sled dogs, the aspiring hockey
star practices slap shots against the garage door, or that those families in the cul-de-sac insist on
an "open door" policy and weekly potlucks.
As the well-informed internet shopper gleans data from websites, the role of the real
estate agent may shift from a home finder to a home guider. Nearly all real estate websites
appear to recognize the important role the realtor plays in the process and lists their names where
possible in the property listing page (Figure 15). All real estate websites contain factual errors or
suspect evaluations; competent realtors can therefore provide the additional, better source of
local knowledge compared to a website.

18

Figure 15 Realtor Links for a Zillow Listing

Given these numerous and varied data prompts homebuyers need to do to perform their
own spatial analysis - what information is most important in the neighborhood search and how
much credence should be given to the datasets themselves? As for the data quality and/or
accuracy question, information that produce red flags for a property may not even be current or
accurate, or can be misinterpreted. Crime data must be up to date and reflect trends, businesses
close, and areas are rezoned for other purposes. Registered sex offenders move and new ones
join the list (and not all offenders register). In examining a crime heat map, a reader may shy
away from areas indicating high crime without realizing that those values are relative and
temporal, and the types of crime included or not included in the map affect its display. For the
prudent shopper, in real estate as in all other things, caveat emptor.
Lastly, the Los Angeles Times developed an online neighborhood map that displays
census data including aggregated statistics for population, ethnicity, income, age, and some
housing data such as rentals (Los Angeles Times 2014). The Los Angeles Times web map project
differs from this thesis web GIS application in that it only provides data that defines and

19

describes neighborhoods, rather than supplying functionality associated with online real estate
searches. The Los Angeles Times is using census tract-level information in an interactive map
format to allow users to understand a fuller picture of a neighborhood in regards only to
demographics.

20

CHAPTER 3: METHODS AND DATA SOURCES
The programming effort in this thesis resulted in a web GIS application that integrates existing
real estate property data with related city and neighborhood data for Anchorage, Alaska into an
interactive local mapping website. The study area was limited to a subset of Anchorage known as
the Anchorage Bowl to minimize the processing time and storage size of the various large
datasets. Anchorage was chosen for several reasons; it has a dedicated GIS user group within the
municipality that creates and serves many appropriate and current datasets that the author has
used before (Municipality of Anchorage 2013). The study area is located in a seismically active
area, therefore, unique local seismic zone data have also been included in the web GIS
application. Anchorage is the author's hometown and where she owns a home, so the author is
familiar with the neighborhoods and issues homebuyers consider and can knowledgably evaluate
the results of this thesis project.
3.1 Study Area
The study area was limited to a section of the Municipality of Anchorage (MOA) that
excludes the surrounding bedroom communities. The city is bounded by ocean to the west and
south, Joint Base Elmendorf-Richardson to the north, and Chugach State Park to the east (Figure
16). This roughly 112 square-mile area is called the Anchorage Bowl.

21

Figure 16 The Anchorage Bowl and Vicinity (Municipality of Anchorage 2013)

3.2 Web Application Development
The research conducted for this project centered on how to display residential areas that
satisfy homebuyers' desired neighborhood features by creating a web GIS application that
displays in a single webpage a detailed map of the study area (the Anchorage Bowl) flanked by a
side panel where users may toggle various layers on and off for a custom display. These map
layers are described Table 1 and Table 2, and consist of schools, trails, parks, seismic zones,
business zones, industry zones, port and airports, major roads, and railroads. Four grouped
selection layers, described in Table 3, are the result of intersections among multiple thematically-
related layers. These resulting web mapping application user selection polygons illustrate the
optimal areas to begin a home search.

22

Table 1 Source Material Information (Municipality of Anchorage 2013)

Layer Theme Source
Original
Format
Date of Collection,
Edit, or Metadata

Crime Hazards APD shapefile
2012 data, layer
created June 2013

Sex Offenders Hazards DPS spreadsheet June 2013
Seismically-Induced
Ground Failure
Susceptibility Hazards MOA GIS shapefile 1979, edited 2005
Floodplain Hazards FEMA Map Service 2014
School Boundaries
Education and
Recreation ASD
keyhole
markup
language 2013
Parks, Trails
Education and
Recreation MOA GIS shapefile
2006, original, but
updated "as needed"
Zoning

Transportation
and Zoning MOA GIS shapefile 2008
Railroad
Transportation
and Zoning MOA GIS shapefile 2008
Roads
Transportation
and Zoning MOA GIS shapefile 1999
Political Boundaries
for State Senate,
State House,
Anchorage
Assembly Districts Political MOA GIS shapefile
Updated after each
election, current as of
2014

Neighborhoods Political MOA GIS shapefile 2008
Anchorage Bowl Political

MOA GIS
shapefile 2008
Properties Housing Gina Bergt, Realtor spreadsheet June 2013

23

Table 2 2012 "Calls For Service" Crimes By Type in the Anchorage Bowl (Morberg 2013)

Crime Type Count
Arson Violent 74
Assault Violent 3530
Burglary Nonviolent 1096
Disorderly Conduct Nonviolent 267
Disturbance Nonviolent 15513
Drugs Nonviolent 1574
Drunk Problem Nonviolent 6296
Fraud Nonviolent 258
Gambling/Prostitution Nonviolent 162
Homicide Violent 15
Liquor Law Violation Nonviolent 470
Misconduct Involving Weapons Nonviolent 1009
Robbery Violent 380
Stolen Vehicle Nonviolent 68
Theft Nonviolent 2472
Vandalism Nonviolent 765

24

Table 3 All Layers That May Be Viewed or Incorporated in a Search
Layer Name Display Included in Search
Trails X Outdoors, Great Outdoors
Parks X Outdoors, Great Outdoors
Neighborhoods X Neighborhoods
Anchorage Assembly District X Anchorage Assembly District
State Senate District X State Senate District
State House District X State House District
Anchorage Bowl X Anchorage Bowl
Alternative School X Alternative School
Charter School X Charter School
Elementary School X Elementary School
Elementary School Boundary X Elementary School Boundary
Middle School X Middle School
Middle School Boundary X Middle School Boundary
High School X High School
High School Boundary X High School Boundary
Crime Heat Map X Crime Heat Map
Seismic Zone X Quiet Zone
Sex Offender X Sex Offender
Flood Plain X Flood Plain
Available Housing X Available Housing
Major Road Great Outdoors, Quiet Zone, Commerce
Railroad Great Outdoors, Quiet Zone
Port Great Outdoors, Quiet Zone
Industry Great Outdoors, Quiet Zone
Business Great Outdoors, Quiet Zone, Commerce

This thesis project was developed primarily using Esri ArcMap 10.2. Layers and data
were initially processed using Esri ArcMap 10.1 and 10.2 to import shapefiles, to convert data to
Google Maps' Keyhole Markup Language Zip (KMZ) layers, and to create datasets from

25

comma-delimited input data. The programming language was hypertext markup language,
revision 5 (HTML 5), using a combination of Esri-created JavaScript v3.7 (3.8) modules (Esri
2013) and other-created JavaScript v3.7 (3.8) with Dojo (Dojo 2014), all using Asynchronous
Module Development (AMD). Dojo acts as a shortcut to JavaScript modules, referencing third
party libraries accessed over the internet, rather than stored on the development server. Dojo
works with AMD to allow the webpage to load quickly, to avoid lag time by only running the
modules as they are called.
All data were projected to NAD83, Alaska State Plane Zone 4. The final web GIS
application was created using datasets published to ArcGIS.com by the author and added to an
Esri Story Map template (Esri 2014) which was modified with programming written in
JavaScript v3.9 and its associated Dojo toolkit (Dojo 2014). Notepad++ was used as the webpage
editing program (Ho 2011). The full process of creating the web application is discussed in
section 3.4.
The flowchart in Figure 17 outlines the process of collecting, editing, and publishing the
data. A collection of relevant data was obtained through municipal and private sources and
organized thematically for processing in ArcMap. The themes chosen cover Hazards, Education
and Recreation, Transportation and Zoning, and Political - groups often-considered elements by
homebuyers who wish to understand a town and its neighborhoods. Hazards includes floodplain
zones, a crime heatmap, locations of registered sex offenders, and seismically induced ground
failure susceptibility polygons. Education and Recreation shows all public schools and MOA
trails and parks. Transportation and Zoning contains several layers including major roads,
railroads, the port and airports, and business and industry zoned areas. Political is a set of
boundaries for the Anchorage Assembly, Alaska Senate, Alaska State House, the names of

26

Figure 17 Flowchart Showing the Find Your Anchorage Neighborhood Creation Process

Determine & Collect
Layers
• Hazards
• Education and Recreation
• Transportation and Zoning
• Political
Configure Data
• Convert or integrate
shapefiles
• Create layers and feature
classes in ArcGIS
• Complete metadata
• Create crime heatmap
• Publish as feature services to
ArcGIS.com
Design Web Application
• Assess usability goals
• Research appropriate map
template
• Create grouped layers for
searches
• Edit attribute data for popup
windows
Investigate ArcGIS scripts
• Investigate Esri sample scripts
• Program in JavaScript and
Dojo against the story map
template
Conduct Analysis
• Create pre-selected user
searches
• Create layers incorporating
related fields for searches
• Publish as feature services to
ArcGIS.com
Complete & Test Website
• Make complete and usable
• Upload all to server
• Plan updates

27

subdivisions in Anchorage, and the study area boundary, the Anchorage Bowl. A final layer
called Available Housing is a sample of housing properties culled from the multiple listing
service (MLS) (Bergt 2013). One of the main objectives of the web GIS application is for users
to be able toggle that layer on and off to view what properties are for sale in their newly chosen
neighborhoods.
In regards to determining the most appropriate software to use in coding this thesis
project, initial research suggested that Google Maps coded with Hypertext Markup Language 5
(HTML5) and JavaScript, Python, MySQL, or Google Fusion Tables (Google 2012) would be
effective to create a simple website application. In contrast to a high level of programming skill,
for example using Java, required to create the desired application using Google products, it was
found that Esri's ArcGIS templates provide pre-coded web GIS applications that can be
customized more easily using JavaScript, Dojo and HTML (Esri 2013). Esri has an active
developers’ website (Esri 2013) to assist web designers and other programmers, by posting
samples of JavaScript and Dojo scripting that can be customized to perform many varied tasks.
Nearly twenty maps using various programming methodologies, scripting languages and
various website examples were created and tested as part of this thesis project. Ultimately, the
ability to create a fully functioning web application incorporating the desired datasets was
successful using Esri's Story Map templates. The latter facilitate focusing on thesis-specific GIS-
created data (Table 2) over detailed programming skills that would have been required to
program the web GIS application from scratch, such as creating and integrating many different
Java or JavaScript applets to allow for the a number of different search options and related map
display.

28

3.3 Source Data
Table 1 identifies all the data created or incorporated for the project. Much of the data
were provided as shapefiles from the Municipality of Anchorage, Information Technology
Department GIS Services (MOA GIS), particularly the zoning, seismic, and political boundaries
(Municipality of Anchorage 2013). The Anchorage School District (ASD) provided school
locations and attendance boundaries in KMZ files (GreatSchools 2013). Crime data covering all
calls for service received in 2012 came from Mr. Bryan Morberg, the GIS Analyst at the
Anchorage Police Department (APD), while the sex offender list of all registered offenders as of
June 2012 was obtained from Citydata.com culled from lists from the Alaska Department of
Public Safety (DPS) (Alaska Department of Public Safety 2013). A random sample of properties
on the market was obtained from Ms. Gina Bergt, a local Realtor with the Jack White Real Estate
office in Anchorage (Jack White Real Estate 2014). The temporal coverage of this data is homes
for sale in June 2013, as an example of a properties dataset that can be loaded directly into this
application. With the exception of the real estate listings, all data covered the geographic extent
of the Municipality of Anchorage rather than just the project limits of the Anchorage Bowl.
These data layers reflect typical homebuyer queries such as schools, crime rates, and
recreation opportunities. Additional features unique to Anchorage were also included,
specifically floodplain zones and seismic designations. Although these elements are also
included in local searches with other online real estate websites, they may not be necessary in all
parts of the country as they are in Anchorage. In some parts of the country, other hazards for
tornadoes, hurricanes, and wildfires may be appropriate for inclusion instead.

29

3.3.1 Source Data for Hazards Theme
There are four datasets in the Hazards theme: crime, sex offenders, seismic areas, and
100-year flood data. Crime data include all reported calls for service (CFS) to the APD in 2012.
These point data were delivered already geocoded in shapefile format and were added to the
geodatabase. The types of crimes are sorted in Table 2 and investigated in the call_type field
within the ArcMap layer attribute table (Figure 18, attribute Call_Type). One specific category is
omitted from this list: sexual assaults (Morberg 2013), whose locations were adjusted by the data
owner to APD headquarters in order to protect the victims. This readdressing would skew the
distribution of crime shown on a heat map and therefore was deemed unsuitable for inclusion in
this web GIS application. The remaining 33,949 (3,999 violent and 29,950 nonviolent) crime
locations are plotted at the 100-level block rather than at their precise location (Figure 19) and
used to create a heat map to show relative quantities of all other crime by location.

30

Figure 18 Example of Crime Types by Date and Address in ArcMap. Note the Sexual
Assault Addresses are Remapped to APD Headquarters.

31

Figure 19 Crime Points Plotted at the 100-Block Level

In general, a heat map is a graduated density surface map created from interpolated
individual points (Morais 2012). In this study, the crime heat map was created through several
steps using ArcMap's Spatial Analyst tool, and provides a continuous surface showing the
geospatial weighting in the number of crimes located throughout a given area. The crime points
vector layer were converted into a raster grid surface via inverse distance weighting (IDW),
where data points closer together are weighted more strongly than those further apart. These data
points were next evaluated to check their statistical significance by attempting to answer this
question: are the crime locations random? A spatial autocorrelation tool, Global Morans I was

32

used, incorporating factors of inverse distance squared and Manhattan distance to determine the
dispersal characteristics of the crimes, depicted as z-scores (Esri 2014). Specifically, inverse
distance squared calculates the influence of one point to another by their distances, in this case a
Manhattan distance that measures along the Anchorage road network. This z-score varies
throughout the Anchorage Bowl, which suggests that the distribution is clustered in the north and
downtown areas where the z-scores are near 10 (indicating intense clustering of crime data), but
dispersed outside those areas (where z-scores are less than 1). More crimes in close proximity to
each other result in greater values in interpolated weighting and in higher z-scores, which display
on the map as brighter and more focused clusters. The resulting heat map data were normalized
into five geographic classes to show the percent of the values relative to the total value of each
geographic class, then assigned a gradient color scheme; higher rated clusters of crime locations
show as darker red while areas of lighter red illustrate lower rated clusters of crime locations
(Figure 20).
A crime heat map is by nature subjective, because by altering the classification scheme or
using different interpolation schemes for given point data different areas of the Anchorage Bowl
would be manipulated to appear more crime-ridden than other areas. As a manual review of the
resulting crime distribution, the published crime heat map was visually compared to the original
point data locations, so that the classification scheme chosen provides the closest visual match
between the cluster distribution of the heat map and the clustering of the original point data. It is
important to consider that the level of both violent and nonviolent crime within the Anchorage
Bowl is relative to other communities in Alaska and outside the state.
Violent and nonviolent crimes were compiled into a single layer in this analysis. The
original consideration was to create separate heat maps to let homeowners weigh the risks of

33

living near nonviolent crime locations compared to locations of crimes involving injury to the
person. On further consideration, by comparing the locations of violent and nonviolent crimes as
shown in Figure 21, it appears that locations of infrequent violent crime are often surrounded by
the locations of more common nonviolent crime. Segregating crime types into multiple heat
maps would not offer significantly different results to warrant displaying more than one map, as
shown in Figure 22.

Figure 20 Classifying and Coloring the Heat Map

34

Figure 21 Anchorage Bowl Violent and Nonviolent Crime in 2012 (Morberg 2013)

35

Figure 22 Anchorage Crime Heatmap

36

The sex offenders layer shows the domicile or employment locations of people convicted
of crimes specified under Alaska Statute 12.63.100, in particular those convicted of sex offenses
or child kidnapping, who are required to register with the state (Alaska Department of Public
Safety 2014). The Anchorage Department of Public Safety maintains the list of registered sex
offenders and provides updates each weekday. (Alaska Department of Public Safety 2013). A
comma-delimited file of the data was obtained from the Anchorage forum of City-Data.com on
June 20, 2013 (City-Data 2013) to provide an example of how this type of data can be integrated
into this application. This original list of 696 names was edited in Microsoft Excel for clarity and
organizational purposes, then imported into Esri ArcMap where 688 addresses were successfully
geocoded using the 10.0 US Streets Geocode Service. The non-matched eight addresses had
house numbers that did not fit any street segment and were thus excluded from this dataset. With
few exceptions for halfway houses, group housing, or rehabilitation centers, the distribution of
addresses extends throughout the Anchorage Bowl (Figure 23).
The sex offender layer was created in ArcMap 10.1 to show just the address, removing all
offender names, crime type and other identifying information. The layer is not included in any
"selection package" search in this project, as no answer may be correct for determining the buffer
width that does not exclude all neighborhoods. This was done not to protect the offender, but
rather to focus on providing only the basic geospatial information relevant to this project. Any
prospective homebuyer will note that a hyperlink to the sex offender registry appears in the
popup window in the web application should the web GIS application user wish to find more
related information from the original data source.

37

Figure 23 Home or Employment Locations of Registered Sex Offenders (Alaska
Department of Public Safety 2013)

38

The Seismically-Induced Ground Failure Susceptibility data were originally created in
1979 and digitized in 2005 by the MOA (Municipality of Anchorage 2013). This layer shows the
probable areas where the ground may slide in the event of a major earthquake. This original layer
is categorized into five classes of ground failure, from "lowest" along the bedrock foothills to
"very-high" near the bluffs and creek mouths. This layer was clipped to the extent of the
Anchorage Bowl, as seen in Figure 24.

39

Figure 24 Seismic Layers in the Anchorage Bowl (Municipality of Anchorage 2013)

40

The FEMA floodplain dataset is an important layer to assess the probability of
catastrophic flooding and used to calculate insurance costs for a property (Federal Emergency
Management Agency 2013). A FEMA-published ArcGIS geoprocessing service (Federal
Emergency Management Agency 2014) showing the 100-year flood zones for the US is
incorporated in this web GIS application, cropped to the extent of the Anchorage Bowl in
ArcMap using the "extract landscape source data" tool (Figure 25). This live geoprocessing tool
was used to create a smaller subset layer for the project geodatabase. The information window
identifies each area by floodplain zone of A, AE, or AH. These areas are all susceptible to a 1-
percent-annual-chance flood and homeowners in those areas may be required to obtain flood
insurance. Specifically, zone A is determined by "approximate methods of analysis", zone AE is
determined by "detailed methods of analysis", and zone AH is in areas where there is "shallow
flooding with constant water-surface elevation ... of 1-3 feet" (Federal Emergency Management
Agency 2014).

41

Figure 25 FEMA 100-Year Flood Plan (Federal Emergency Management Agency 2014)

42

3.3.2 Source Data for Education and Recreation Theme
All public school data were provided by ASD in a zipped KML or KMZ format to
include point data for elementary, middle, and high schools and polygon layers for their
attendance boundaries, plus point data for charter and alternative schools (Figure 26)
(GreatSchools 2013). The data were imported to ArcMap, clipped to the Anchorage Bowl study
area limits and color-coded to show the school levels. All schools and boundaries are displayed
in the web application. No private schools were added nor were post-secondary or trade schools
(GreatSchools 2013).
Also included in the Education and Recreation theme are two layers, city parks and city
trails. These shapefiles were imported to ArcMap and clipped to the study area extent (Figure
27). These only show areas created and maintained by the MOA - no state or federal land is
included.

43

Figure 26 All Anchorage School District Schools and School District Boundaries
(GreatSchools 2013)

44

Figure 27 MOA Trails and Parks (Municipality of Anchorage 2013)

45

3.3.3 Source Data for Political Theme
The polygon boundaries for political districts, both at the municipal and state level are
included in the Political theme along with a boundary polygon file for the Anchorage
neighborhoods (Figure 28). The polygon outlining the Anchorage Bowl study area is also
included in this frame. Popup windows in the web application show both the district
identification and the current office holders as updated by the MOA GIS in the district polygons
(Municipality of Anchorage 2013). Alaska has only one United States (US) Representative;
therefore, a polygon outlining US House districts is unnecessary.

46

Figure 28 Political Theme (Municipality of Anchorage 2013)

47

3.3.4 Source Data for Transportation and Zoning Theme
The Transportation and Zoning theme contains shapefiles that are buffered and used to
find residential polygons outside these buffered areas. The most manipulation required to prepare
data for this project involved many of these zoning shapefiles. The railroad file was clipped to
the Anchorage Bowl layer, and the major roads were identified from the street segments
shapefile. The zoning shapefile was queried by zoning description (Municipality of Anchorage
2013) to derive several feature classes: business polygons, industry polygons, and the port and
airports in Anchorage. A fourth feature class, called All Housing, was also created to show
residential polygons that are stored in the Housing theme (Figure 29) Several of the zoning codes
were generic and could refer to more than one district, for instance R-11 and PC (Figure 30).
Using local knowledge against Anchorage-area aerial and satellite images or by actually visiting
the locations, the author investigated those polygons and individually assigned them to the
zoning district that best fit their current use.

48

Figure 29 Transportation and Zoning Theme (adapted from (Municipality of Anchorage
2013)

49

Figure 30 MOA Zoning Codes By District Type from (Municipality of Anchorage 2013)

3.3.5 Source Data for Housing Theme
This housing theme is composed of two data layers: one showing available properties and
the other the result of the zoning layer query by residential type (Figure 31). The available
properties layer is a random sample of units on the market in the Anchorage Bowl as of June
2013; it is used solely as an example dataset to illustrate real estate availability after the
neighborhood search has been completed. As previously stated, this layer was created from a
query of the Anchorage MLS website conducted by a local Realtor (Bergt 2013). The properties
of interest were saved and organized in Microsoft Excel.Fields were created for price, bedroom
and bathroom quantities, address, square footage, lot size, garage or carport stall slots, and listing
office. The MLS number supplied was reassigned to a simple sequential object ID value for
consistency (Figure 32). The Excel file was imported into ArcMap and geocoded using the 10.0
US Streets Geocode Service (ArcGIS Online, 2013). Of the 140 properties, only seven did not
match an address; these properties were on lots with new streets that had not yet been added to
the service. Thus those properties had to be excluded from the resulting dataset.

50

Figure 31 Housing Theme (Municipality of Anchorage 2013)

51

Figure 32 Excerpt from the Available Housing Attribute Table

The residential zoning layer was extracted from the zoning shapefile and saved to be used
as a clip feature against the selectable layers. It is these resulting polygons that reflect the
optimal neighborhoods for the subsequent residential search, while the residential zoning layer
itself is not displayed in the map. Two other MOA shapefiles, neighborhoods and subdivisions
derived from the MOA GIS landuse shapefile could be logical choices, but neither are good
choices. The neighborhoods layer polygons are too large and user searches would not result in a
very narrow or specific area. Conversely, selecting areas based on the small subdivision
polygons would yield too many options (properties) for users as the polygons are much smaller

52

and successful searches would be widely scattered across the study area. A comparison of the
two sizes is shown in Figure 33.
The best choice that reflects both the most up-to-date residential zoning and large
polygons is found in the All Housing layer derived from the Zoning shapefile. While the
neighborhoods layer was not used as a clip layer, it is still valuable to show on the map to aid
realtors and homebuyers with the names of the local areas.

53

Figure 33 A Size Comparison of Neighborhood Versus Subdivision Layers

54

3.4 Data Configuration in the Find Your Anchorage Neighborhood Application
In the web GIS application the toggled (on or off) layers and the selected database query
results are displayed as polygons: (business, industry, and port buffers at distances of 1/4-mile
and 1/2-mile, parks, school attendance, seismic, political boundaries), points (schools,
properties), lines (roads, railroads, and trails), or interval maps (crime heat maps). The layers are
grouped by common themes in ArcMap as per the subsection headings in this chapter, and then
published as Esri ArcMap feature services that are incorporated into the web GIS application
coding.

Figure 34 Data Preparation Tasks Performed on Trails, Parks, and Seismic Layers

All data were clipped to the extent of the project study area boundary using the
Anchorage Bowl layer. Additional processing was required for layers used to intersect with the
All Housing layer and used for one of four of the selections packages created in this project. The
Trails
Clip
All Housing

1000-foot
buffer to
include
Processed
Trails
Processed
Parks
Parks
Select
From
1000-foot
buffer to
include
Seismic
Select
From Zones 3, 4, 5
Processed
Seismic

55

flowcharts in Figure 34 and Figure 35 Layers Created From Zoning Layerillustrate the ArcMap
query processes that created these selectable layers. Most of those searchable layers are derived
buffered layers altered with regard to residential zoning parcels. For instance, homebuyers may
want to be within walking distance of parks and trails and further from major roads and industrial
areas; therefore the buffered layers were set to include housing options within 1000 feet of trails
or parks, and exclude residential areas that fall within either a quarter-mile or half-mile of
business parks and airports. All of these resulting parcels were then clipped against the
residential polygons of the All Housing layer since homebuyers cannot purchase or build homes
not zoned for residential use. Because these buffer distance decisions were pre-determined, the
buffered layers could be generated server side and combined as static data layer options to be
chosen by the homebuyer, rather than generated on-the-fly on the client side which would have
greatly slowed map rendering. No buffered layer, however, can be displayed in the map except
as part of a search. This helps to minimize clutter in the resulting web GIS map, and fits in with
the established searches that have been created. Pertinent layers can be toggled on and off within
the map display using code adapted from the original template coding. A complete list of all
layers is in Table 2.

56

Figure 35 Layers Created From Zoning Layer

Once all the data layers are created and processed in ArcMap, selected layers can be
integrated into a proper search function in the web GIS application. As previously mentioned,
there are myriad options that users may want to consider when performing a search for potential
homes: proximity to specific schools, are they within walking distance of offices, locations in
"safe" (crime) areas, or combinations thereof. Due to complications discussed in Chapter 4, the
author elected to make "selection package" searches using the processed layers created as
illustrated in Figure 34 and Figure 35, plus the transportation layers or port/airport, rail, and
major roads. Table 4 lists the four searches and the layers employed in the design of the web

57

map. Outdoors is for those homebuyers who want to be near the extensive trail or parks system,
while Great Outdoors adds the exclusion zones of business, industry, roads, rail, and ports, for
those who want to "get away from it all." For those who crave peace and quiet, the exclusion of
business, industry, roads, rail, and ports is added to the more seismically-stable zones in the
Quiet Zone. The fourth search assumes a homeowner who wants or needs to be closer to the
city's amenities, for either work or commerce. Commerce includes commercial or business
parcels (properties) within 1/2-mile of those zones and within 1/2-mile of major roads. All of
these “selection packages” created for the user are Esri feature services published server-side,
rather than client-based, which saves map-rendering time as previously mentioned.

Table 4 "Selection Package" Search Feature Services
Search Name Includes Layers Buffered Distance
Outdoors Trails 1000 feet inclusion
Parks 1000 feet inclusion
Great Outdoors Trails 1000 feet inclusion
Parks 1000 feet inclusion
Port 1/2-mile exclusion
Major Roads 1/4-mile exclusion
Rail 1/2-mile exclusion
Industry 1/2-mile exclusion
Business 1/2-mile exclusion
Quiet Zone Port 1/2-mile exclusion
Major Roads 1/4-mile exclusion
Rail 1/2-mile exclusion
Industry 1/2-mile exclusion
Business 1/2-mile exclusion
Seismic Zones 3, 4, 5
Commerce Business 1/2-mile inclusion
Major Road 1/2-mile inclusion

58

3.5 The Web-Based Design Application Development
The software and scripting samples most helpful in creating the website were readily
available from Esri, specifically, ArcMap 10.1 and 10.2, Esri Story Maps templates, and
ArcGIS.com. Notepad++, a free online editor with a clean and intuitive design, was relatively
easy to use to edit the various web GIS application files (Ho 2011).
As previously mentioned, all data were edited using ArcMap 10.1 and 10.2. Common
layers were grouped into themes and then published directly to ArcGIS.com, which provides
unique URL addresses that are entered directly to a web GIS application. As all the layers within
a given theme are published together, each layer has a unique value at the end of the frame’s
URL that allows each layer (representational state transfer, or REST endpoint) to be referenced
in the web application, as shown in Figure 36. For example, the Hazards theme includes four
layers, and the unique value representing Seismic is the number 0 at the end of the URL.

Figure 36 Seismic Layer URL

59

3.5.1 Web Application Planning
Using the data layers completed and published to ArcGIS.com, many iterations of the
final map were attempted using Esri tools and JavaScripting code to find a method that
successfully integrated all of the data, JavaScript modules and other scripts into a single working
website. The original goal of the this project was to show a map of the Anchorage Bowl with
additional interactive features flanking the map on either side of the web map, in a single web
GIS application. This more complex design was later abandoned due to numerous coding
difficulties integrating code snippets and modules in several scripting languages. The
ArcGIS.com template model was ultimately adopted as described in subsequent sections of this
thesis.
This section documents the design and testing of the first web application development
plan. On one side would be a toggable list of all the created map layers so that the user could
acquaint himself with the town. On the other side would be a subset of those layers from which
the user could select items pertinent to a custom search. Those selected elements would only
display where they intersected each other, the result being the user's optimal areas of town. An
additional program element considered included network analyst, which can plot commute times
between homes and work sites that ideally could be integrated into each search. Once the search
was complete, the user could then toggle the AvailableHousing layer to see which properties
were in those optimal areas. Popup windows would give information about the homes, including
the realtor contact information.
The trial development website was served via the Internet Information Services (IIS) and
accesses the data published on the ArcGIS.com server. A link to this thesis is also available in
the About section of the final web application.

60

3.5.2 Web Application Development Efforts
According to Zed Shaw, "programmers are lazy" and will seek to use code programmed
by others (Shaw 2010). This is not to suggest programmers are thieves but that many of the
scripts used in numerous applications are not unique. The robust Esri Developers Forum (Esri
2013) and the user forum from Stack-Overflow (Stack-Overflow 2013) contain many scripting
examples. The Esri ArcGIS API for JavaScript page contains many examples for developers to
try out in a "sandbox" environment; successful edits can easily be copied to the web page
programming (Esri 2013). While searches for web code to address specific problems may be
hard to find, it was often not difficult to find someone to share code they have already created for
an existing and usually published (publically available) application. The challenge then was to
fully understand how to modify and incorporate the code into the existing program without
adversely affecting other sections.
Many versions of the application were attempted, with each revision developing or
abandoning a subset of the overall code while keeping the main goals in mind. Usually this
meant establishing good coding habits by creating separate web pages to alter just the new code
or manipulate sample programs; if that ran successfully, those changes could be introduced into
the main webpage. Not all programming efforts successfully integrated, for example as stated
previously, when trying to integrate code in different languages, different versions and from
different sources, the customization was extremely difficult. As a result, significant revisions to
the layout and coding had to be made, and the web application goals were altered to be more
realistic for time and programming constraints.
While the web forums and conversations with Esri programmers were often beneficial in
learning the processes of programming the map, particularly in how the scripts were constructed,

61

the author found the successful creation of a robust web application by modifying these
numerous scripts a challenging task that eventually overshadowed the learned GIS processes. For
that reason, the author abandoned the initial program and instead sought a published template
from Esri's Story Maps collection (Esri 2014). These customizable templates allow the
programmer to add created and published GIS maps and data. While there was still a significant
amount of programming required to edit the template to satisfy the overall goal of this thesis, it
was no longer as onerous a programming task. As a bonus, the template already included useful
default design amenities such as an index map, a map/image toggle button, a return-to-extent
button, and zoom levels.
3.6 Final Web Map Programming
Using the Story Map template meant altering some of the design elements of the web
application, most notably the user's ability to create his or her own search parameters. Only a few
template options closely matched the original design of the web application, but one from the
repository of the website GitHub (GitHub 2014) proved to be the most useful. The flowchart
shown in Figure 37outlines the overall methodology for customizing the ArcGIS.com template
to produce the final, successful web application.

Figure 37 High Level Flowchart of Web GIS Application Creation Steps
ArcMap
• Edited, Symbolized, Added Metadata
• Published to ArcGIS.com Server
ArcGIS
• Created One Map of All Layers
• Configured Popup Windows
Story Map
• Added Via REST Endpoints All Layers as Code in Template JavaScript Files
• Edited CSS Files to Change Look of App

62

The key elements desired are already included in the original GitHub-based template: a
means to let the user select the layers to display, display a map legend, and interact with
descriptive popup windows. As seen in the template sample in Figure 38, the map is uncluttered
and easy for the user to understand and manipulate due to the inclusion of extensive
programming libraries of associated style sheets, JavaScript files, and configuration codes.

Figure 38 Public Information Template
Source: (GitHub 2014)

63

A benefit to using this particular template is that it allows the developer to directly insert
a completed map published via ArcGIS.com by adding it to the web application using its unique
webmap ID, as shown in Figure 39. Many tasks related to finalizing the map can also be
completed using functionality provided in ArcGIS.com, when a developer is logged in and
working on a web map or web GIS application. A side drawer, an expandable web page panel on
the left side of the application including three separate tabs, lets users toggle layers to be
displayed when opening the map and the basemap is chosen to suit the needs of the program.
Bookmarks were created to pinpoint sectors in detail, and popup menus were configured for
greater clarity. The layer displays were also altered to show only within a set zoom level or be
made more transparent so they reduce visual clutter. Any modifications saved in the ArcGIS.com
hosted web map automatically show up in the live web application that the user can access.
The main problem with hard coding the web GIS application was the inability to
successfully create and display the results of searches. While no Esri Story Map template has
been created yet that solves this problem, there is a viable work-around. Rather than attempt to
set up a cumbersome search process for all layers, the author created four processes that can be
called up by the user of the web GIS application. These searches as defined in Table 3 were
added to the Story Map template where the user can toggle them on and off. An example of a
Great Outdoors selection is shown in Figure 40, where popup windows provide descriptive
information.

64

Figure 39 ArcGIS.com Map With Webmap ID

65

Figure 40 Great Outdoors Selection with Descriptive Popup Window

One additional feature in this template is a swipe function that displays a separate layer as
the user scrolls a vertical line back and forth (horizontally) across the page. In the case of this
web GIS application, the All Housing features are revealed using a vertical swipe bar as seen in
Figure 41. Generally, a swipe bar is employed to show differences between map layers in image
format, but in this application, the swipe function allows the user to quickly drag the All Housing
layer features over an area to display properties in the user’s neighborhood area of interest.

66

Figure 41 The Swipe Display Option in the Web Application

67

CHAPTER 4: RESULTS
The final web application developed as part of this thesis is available at http://591-
jndowlin.usc.edu/storymap/ThisTemplate/index.html. This application provides the user an
opportunity to view layers that help orient him or her within the city of Anchorage, and several
search packages filter residential zones based on the user’s need to either remove him or herself
or include him or herself in the activities of the Anchorage Bowl. Once a user has identified his
or her ideal neighborhoods, the user can display available properties by swiping that layer over
the display to begin a home search.
4.1 A Tour of the Find Your Anchorage Neighborhood Application
The web GIS application usage is straightforward in its approach. An "About" screen
appears first to explain the application and instruct the user how to select layers and choose one
of the "selection packages." Three tabs in the left side drawer give the user flexibility to select
layers to view, conduct searches in the map, visualize the layer symbology, and read metadata.
The first open tab is the "About" view (Figure 42) which links the user to the metadata
and instructions page (Figure 43). This tab also houses the bookmarks that link to a selected
neighborhood. This useful feature is important in displaying larger layers that cannot fully render
in the map but will display completely at a larger zoom level. An unusual feature is the current
date, useful indicate when the map was last edited. Few websites add the last edited date, but
should do so to show that the website is not static.
The "Legend" tab shows the layers that have been toggled in the "Layers" tab and
displays their icons and colors as created in ArcMap. By default, the Neighborhoods layer
appears when the map is opened, as is the Available Housing that is visible only in the swipe
function. All other layers are toggled as desired. All layers display at all zoom levels, except

68

those that approach or exceed the limit of 3000 rows that then render slowly or not at all. In this
web GIS application, the large Trails layer and Commerce search layer may require the user to
zoom in or select a large-scale bookmark to avoid long render times. As in ArcMap, layers at the
top of the list may overlap those below them, including the selection layers at the very bottom.
Therefore, these selection layers are best viewed with the other layers turned off.

Figure 42 The Three Tabs in the Application Expanded

69

Figure 43 Partial Metadata for the Find Your Anchorage Neighborhood Web GIS
Application

Two more interactive features are set in the original template. The website user can enter
an address or business name into the search window to quickly find and zoom to that spot. That
search function will allow the user to type in any street or business, even if it is not in Anchorage
(Figure 44). Nearly all Story Map templates come with a way to broadcast via social media. Map
users may embed the map with the provided HTML code, or share it through email or other
media sites, as seen in Figure 45.

70

Figure 44 One of Several 1600 Pennsylvania Avenues

Figure 45 Social Media Options

71

4.2 Challenges Encountered
Several significant impediments affected the success of this project. These impediments
are broken into issues concerning the Esri ArcGIS suite and how it is integrated with SQL
Server, and the web programming using HTML and JavaScript. A great deal of database design
and permissions handling is required to implement SQL Server as a backend for ArcGIS Server.
The web page scripting was challenging primarily because the author’s education and prior
experience did not include additional computer science background outside of scripting learned
in recent GIS programming coursework. It was these issues that took up most of the time and
effort in the latter part of this project, as described below.
4.2.1 Esri ArcGIS and Microsoft SQL Server
Despite being a logical choice to use in this project due to increased familiarity and
practice with robust software, Esri's ArcGIS suite is not a user-friendly product. A number of
problems occurred with data permissions, delayed registering, enterprise geodatabases and
publishing feature services necessary to create the web GIS application. Incredibly patient Esri
technical support managed to straighten out the many issues, mostly by establishing SQL Server
accounts and turning properties settings for "administrator" rather than "publisher" privileges.
This corrected authentication errors that prevented accessing and saving data in geodatabases
within ArcMap, and from publishing to the ArcGIS.com server.
An unexpected consequence was ownership of files within the geodatabase. Two
authentication paths are available when using a database connection - the operating system
authentication and the database authentication. Whenever a schema is established, the data are
created and saved to a database; that process dictates the ownership of the data. The operating
system does not require a password and therefore cannot be used when the data will be published

72

to another server. Many of the layers in the project were saved that way. Once the decision to
publish the data was made, the permissions had to be changed and the database connection had
to then use a password-enabled database authentication. Data saved to the geodatabase after that
point were saved in that way and all other data were pointed to the database authentication
scheme for publication. Any data files that required edits, however, had to point to the owner
schema, before any changes could be made. This was both time-consuming and confusing. Any
researcher who plans to create and publish datasets using SQL Server connections should be well
versed in its theory and application, if only to avoid such an amalgamation of schema within the
same database.
The project originally began in ArcMap 10.1, but the software became "buggy" in that it
slowed processes and often prevented tools from working. Eventually, the project migrated from
10.1 to 10.2, and once the SQL Server connections were reestablished most of the problems with
the program disappeared. One problem that plagued the project throughout was space limitations
to the hard drive. Interestingly, that error manifested itself in different ways. Processes that
worked perfectly previously were now unworkable, illustrated by permissions errors or locks to
the data. After several failed attempts to complete a task, the program would flash a space limit
screen (Figure 46). Deleting unneeded data files and programs from the virtual machine was the
only way to eke out room to complete processes.

73

Figure 46 Insufficient Space Allocation Error in ArcMap

These challenges highlighted problems with the Esri Help file: while extensive and very
current, it tends more toward the theory behind a task rather than the steps needed to accomplish
it. The theory description relies heavily on jargon and detailed GIS concepts and by providing
myriad links to help webpages that take hours to root through to a solution. In the instances listed
above, only with the help of Esri Support could the problems be efficiently fixed.

4.2.2 HTML and JavaScript
Web programming is an intricate and time-consuming task, and combining HTML and
JavaScripting with Esri-created geospatial data can overwhelm a novice programmer. Esri's
Developers Forum for JavaScript applications has many small code samples for programmers to
add to their own programs, and allows developers to use a "sandbox" to modify the scripts in

74

isolation from their main page (Esri 2013). A few of the scripts appeared useful enough in this
thesis to warrant investigation, but none directly solved troubles that plagued the first iteration of
the thesis web programming effort.
Many JavaScript and Dojo scripts from other websites were researched, especially those
associated with geospatial programs or in help forums of StackOverflow (Stack-Overflow 2013).
Curiously, because JavaScript versions changed often, certain compatibility issues required many
hours of tinkering to correctly load the map and implement new elements. Map load functions,
for instance, conflicted with instructions in other parts of the file, which meant the webpage
would not render correctly. This problem most likely was a result of the author who did not fully
understand the coding structure and how it related to other modules within the application. A
simple workaround in these cases was to add additional lines in the JavaScript heading to
recognize all JavaScript libraries.
A main conundrum in the web GIS application programming was the drop down menus
from which the user could select multiple queries. Most existing real estate websites do allow the
homebuyer to choose from drop down menus and display the results on the map (Figure 47). In
many cases, these websites user Google maps and API functions to help produce their product
(McConnathy 2013), citing customization ease. The Esri JavaScript code samples had no
offerings for drop down menus and the samples for database queries did not provide a means to
display the results in the map. This query does exist in ArcMap where users can select rows to
display from a layer's attribute table, so it is confusing that it does not exist in other Esri code.
New samples are created often, and such a process may yet be added to Esri templates, which
when combined with the geospatial applications in ArcMap could be integrated to produce a
more powerful web application.

75

Figure 47 Drop Down Menu Option in Trulia.com

One of the final challenges in the coding scheme was to incorporate a swipe
function in the Story Map template. The author studied the many JavaScript files
included in the template to determine where to add the All Housing layer link. A series of
email exchanges with the template author (Driscoll 2014) provided the answer. Each
published layer is referenced by a unique text ID that is different from its ID in the REST
services folder as shown in Figure 36. This ID may be found using developer debugging

76

tools such as Firebug; the text ID for the All Housing layer is "Housing_7118" (Figure
48) which is added to the defaults.js file in the template.

Figure 48 Finding the Layer Text ID Using Firebug

77

CHAPTER 5: CONCLUSIONS
A web application that emphasizes neighborhood features before home amenities is a viable
approach to home buying. While many real estate websites add this information after the user
clicks on properties, other websites focus entirely on neighborhoods at the expense of property
listings. The common factor for these websites is the ability to conduct realty research online,
both with and without a realtor's aid. For this reason, both types of websites must be accurate and
current and respond to evolving programming schemes. Future updates or additions to this
dataset can be easily applied, as described in the future maintenance discussion in this chapter of
this thesis.
5.1 The Future of Real Estate Search Websites
As more people use the internet to research tasks that previously were only performed by
industry professionals, the proliferation of specific websites that show real estate listings will
increase. Beyond showing an assortment of photographs and producing a checklist of home
amenities, realtors and real estate web programmers will focus on providing a unique
experiences for prospective homebuyers. They could potentially describe a neighborhood in full
by incorporating census data, linking news reports, analyzing economic trends, and integrating
social media posts. Greater search options will lead to detailed home and neighborhood
descriptions to populate websites' search databases. Home searches will be more than paper
flyers retrieved from a For Sale sign yard box, but instead be an invaluable tool for both realtor
and homebuyer so that both parties can understand what aspects of a neighborhood and home are
attainable. In many cases, real estate websites cannot supplant the expertise and experience of a
realtor, but will instead enhance the process.

78

5.2 The Ever Expanding Future of Web GIS Map Programming
Esri's JavaScripting samples (Esri 2013) is an example of ever-expanding geo-
programming interfaces. JavaScript and HTML code samples currently available are often edited
by developers to incorporate recent programming improvements. Because of these
advancements, someone who understands the changeable nature of web GIS map application
programming should continuously monitor the relevant programming language to ensure that
their own websites continue to display and function properly.
There are many elements incorporated in this web GIS application to provide the basic
information required by a residential real estate client to search for a new home, plus some
functions that improve the usability or display of the resulting maps. The great benefit of using
templates provided by Esri's Story Maps and GitHub enable a more direct application of GIS-
derived maps into a publicly available web application. Sites such as these add new scripting
codes and templates frequently to aid less competent web programmers who deliver the results of
GIS effort. For example, the template used in this project was created ten days before it was
selected.
Esri is not unique in providing tools to help real estate web designers. Google APIs are
integral to real estate websites such as Zillow and Trulia, and these larger companies employ
many programmers to develop unique code in response their company's needs and even enter
into partnerships with other companies to deliver real estate information (Zillow 2014). Many of
these sites do not provide any neighborhood information, particularly the locally owned real
estate offices, but as programming script become more prevalent or if homebuyers insist on a full
picture of the area, this may change.

79

5.3 Future Work on the Find Your Anchorage Neighborhood Application
This web GIS application was designed as an interface to a dynamic real estate search
page, therefore data maintenance is required. Crime and sex offender data are the most fluid and
that information should be updated often. The lists of office holders in the municipal and state
political offices may change after each election through the shapefiles presented by the MOA
and should be updated in the database. Zoning and school shapefiles should also be periodically
inspected for edits to location or classification. While it was only a secondary feature in this
project, property lists from realtors and MLS databases should be complete and current, perhaps
via live feeds from original database sources. Future work should investigate the constantly
changing elements of web design and database management and add such to the website if it is
deemed to improve the user experience. Because this application was created using Esri
software, it is easiest to continue to host the map using ArcGIS.com, a solution that can be
readily incorporated into a company's existing website. A national-based website would require
data to be added to this application from many other communities, so the time frame to add all
available data would be dependent on the amount of information that could be collected and
integrated into the existing application.
The original design for this web GIS application was ambitious and was slowly scaled
back in response to programming challenges or data acquisition issues. Some datasets that would
have been good to include were unavailable from MOA, or only offered as printed maps, such as
tax roles, municipal bus routes, private schools, and emergency service locations. Digitizing
these layers, while very useful in a home search, would be time consuming, and except for the
bus route data, would not have been applicable to any of the "selection packages" searches in this

80

project. Rather, these data could be included in later work to allow additional advanced queries
within a more custom and specialized search tool.

Figure 49 Crime Mapping Display From Trulia.com

Several data layers were added as information-only layers, namely the flood zone, crime,
and schools layers. It would be very useful to allow homebuyers to conduct searches on school
names or indicate "safer" areas of town. Currently, no real estate website offers all these search
options. Figure 49 shows a robust crime layer from Trulia where users see recently reported
crimes used to create updated heat maps, look at trends, and toggle various crime types. The
homebuyer, however, cannot easily choose to see homes for sale only in the "green" areas. As
discussed in chapter 3, heat map displays can be subjectively interpreted. Census data was not
included in this project, but could be useful in future work if treated sensibly. The author

81

considered incorporating neighborhood information about age, income, ethnicity, and housing
type. Such data would help paint a more vivid picture of a given neighborhood, but has a real
potential to be misconstrued. Realtors are wary of or even prohibited from discussing sensitive
Census data for fear of treading on Fair Housing Laws (Wiggin 2014). These concerns are
outside the scope of the thesis and are not discussed further herein.
One final suggestion for future work is to create a mobile application handy for potential
homebuyers as they travel to their optimal neighborhoods to begin directed home searches. The
effort needed to develop a clean application for smaller devices may be significant, but as more
users rely on mobile devices, its development is imperative for a real estate web site to maintain
its competitive edge.

82

REFERENCES
Alaska Department of Public Safety. 2014. Chapter 12.63. Registration of Sex Offenders.
http://www.legis.state.ak.us/basis/folioproxy.asp?url=http://wwwjnu01.legis.state.ak.us/c
gi-bin/folioisa.dll/stattx11/query=[group+section1263100!3A]/doc/{@1}/hits_only
(last accessed 30 May 2014).
——. Sex Offender/Child Kidnapper Registry. 2013. Alaska Department of Public Safety.
http://dps.alaska.gov/sorweb/sorweb.aspx (last accessed 20 June 2013).
BAIR Analytics. 2011. RaidsOnline: Crime mapping, email reports and tips for the public.
http://raidsonline.com (last accessed 10 September 2014).
Benjamin, J. D., P. Chinloy, G. D. Jud, and D. T. Winkler. 2005.Technology and Real Estate
Brokerage Firm Financial Performance. The Journal of Real Estate Research 27: 409-
426.
Bergt, G, interview by the author. 2013. Realtor, Jack White Real Estate.
Choziak, E. R. 2003. Method of Creating and Distributing Real Estate Marketing Packages
Utilizing the Internet. United States of America Patent 0130928.
City-Data. 2013. Registered Sex Offenders in Anchorage, Alaska. City-Data.
http://www.city-data.com/so/so-Anchorage-Alaska.html (last accessed 20 June 2013).
Dojo. 2014. Dojo. http://dojotoolkit.org/ (last accessed 8 July 2013).
Driscoll, M., interview by the author. 2014. Programmer, Esri and GitHub.
Eraker, D., A. Dougherty, E. Smith, and S. Eraker. 2005. Web-Based Real Estate Mapping
System. United States of America Patent 0288957.
Esri. 2013. ArcGIS API for JavaScript. Esri. https://developers.arcgis.com/en/javascript/ (last
accessed 26 June 2013).

83

——. 2013. Mapping APIs and Services for Developers. Esri. http://www.esri.com/getting-
started/developers (last accessed 26 June 2013).
——. 2014. Spatial Autocorrelation (Morans I) (Spatial Statistics). ArcGIS Resources.
http://resources.arcgis.com/en/help/main/10.2/index.html#//005p0000000n000000
(last accessed 26 August 2014).
——. 2014. What is a Story Map? http://storymaps.arcgis.com/en/articles/what-is-a-story-map/
(last accessed 18 April 2014).
Federal Emergency Management Agency. FEMA 100-Year Flood Zones in the USA
(MapServer). 2014. https://maps3.arcgisonline.com/ArcGIS/rest/services/A-
16/FEMA_100-Year_Flood_Zones_in_the_USA/MapServer (last accessed 20 July
2014).
——. 2013. Flood Zones. http://www.fema.gov/floodplain-management/flood-zones (last
accessed 19 May 2014).
Florance, A. C., R. Evatt, D. Hofman, J. Choi. 2007. System and Method for Accessing
Geographic-based Data. United States of America Patent 7174301.
GitHub. 2014. Esri - public-information-map-template-js. https://github.com/Esri/public-
information-map-template-js (last accessed 25 April 2014).
Google. 2012. Google Fusion Tables. http://www.google.com/drive/start/apps.html (last
accessed 29 May 2012).
GreatSchools. 2013. Anchorage School District Schools. GreatSchools.
http://www.greatschools.org/search/search.page?q=anchorage%20school%20district&sta
te=AK (last accessed 30 June 2013).

84

Hagey, P. 2014. Zillow, Trulia, realtor.com now capturing 1 in 2 desktop visits in real estate
category. Inman News. http://www.inman.com/2014/03/03/zillow-trulia-realtor-com-
now-capturing-1-in-3-desktop-visits-in-real-estate-category/ (last accessed 11 June 2014).
Hancock, S. L. 2003. Geographic Location Multiple Listing Service Identifier and Method of
Assigning and Using the Same. United States of America Patent 6597983 B2.
Ho, D. Notepad++. 2011. http://notepad-plus-plus.org/ (last accessed 14 March 2012).
Hoak, A. 2010. After buying home, plan for expenses. Los Angeles Times.
http://articles.latimes.com/2010/jan/01/business/la-fi-home-costs1-2010jan01 (last
accessed 4 March 2014).
Hoffman, K., J. Anzalone, and C. Cormack. 2003. Automated Method and System for Managing
and/or Transferring Real Estate Information. United States of America Patent 0220807.
Homefacts. 2014. Homefacts. http://www.homefacts.com (last accessed 11 June 2014).
Jack White Real Estate. 2014. Gina Bergt, Jack White Real Estate.
http://ginabergt.jackwhite.com/ (last accessed 26 June 2013).
Light, J. 2014. In Zillow-Trulia Deal, Making Room for Brokers. Wall Street Journal.
http://online.wsj.com/articles/zillow-to-buy-real-estate-rival-trulia-1406551333 (last
accessed 31 July 2014).
Los Angeles Times. 2014. Mapping L. A. Los Angeles Times.
http://maps.latimes.com/neighborhoods (last accessed 1 September 2014).
McConnathy, J. 2013. Google Maps Helps Trulia Put Some Heart Ito Home Buying. Google
Enterprise. http://googleenterprise.blogspot.com/search?q=Trulia (last accessed 2 June
2014).

85

Morais, C. D. 2012. Heat Maps in GIS. http://www.gislounge.com/heat-maps-in-gis/ (last
accessed 30 May 2014).
Morberg, B., interview by the author. 2013. Anchorage Police Department Crime Analyst (24
June 2013).
Multiple Listing Service. 2012. MLS. http://www.mls.com (last accessed 28 November 2012).
Municipality of Anchorage. 2013. GIS Data & Document Downloads. Municipality of
Anchorage. http://www.muni.org/Departments/it/GIS2/Pages/MOAGISData.aspx (last
accessed 16 June 2013).
——. 2013. Municipality of Anchorage Planning Department. Planning Division.
http://munimaps.muni.org/planning/zoningsummarychart.pdf (last accessed 11 August
2013).
——. 2013. Zoning Districts. Municipality of Anchorage.
http://www.muni.org/Departments/OCPD/Planning/zoning/Pages/ZoningDistricts.aspx
(last accessed 28 June 2013).
National Association of Realtors. 2013. 2013 Profile of Home Buyers and Sellers. National
Association of Realtors. http://www.realtor.org/reports/highlights-from-the-2013-profile-
of-home-buyers-and-sellers (last accessed 3 June 2014).
——. 2014. NAR Website Traffic Stats. Realtor.org.
http://www.realtor.org/sites/default/files/reports/2014/nar-website-traffic-stats-march-
2014.pdf (last accessed 19 June 2014).
NeighborhoodScout. 2014. NeighborhoodScout. http://www.neighborhoodscout.com (last
accessed 5 June 2014).
Redfin. 2012. http://redfin.com (last accessed 30 November 2012).

86

Shaw, Z. A. 2010. The Command Line Crash Course.
http://learncodethehardway.org/cli/book/cli-crash-course.html (last accessed July 2012).
Stack-Overflow. 2013. http://stack-overflow.com (last accessed 18 September 2013).
Trefethen, S. 2013. Online real estate sites transforming industry. Real Estate Weekly.
http://bi.galegroup.com.libproxy.usc.edu/essentials/article/GALE|A323659119/4c9137a0
66b029a6ce1d9bf73df5a31f?u=usocal_main (last accessed 18 June 2014).
Trulia. 2012. http://www.trulia.com (last accessed 28 November 2012).
Wiggin, T. 2014. Steering 2.0? Data may undermine fair housing laws. Inman News.
http://www.inman.com/2014/04/29/steering-2-0-data-may-undermine-fair-housing-laws/
(last accessed 11 June 2014).
Zillow. 2014. Zillow API Network Examples. Zillow.
http://www.zillow.com/howto/api/APIPartners.htm (last accessed 3 June 2014).

Abstract (if available)

Abstract Online geospatial data are evident in many websites, covering a variety of interests such as route planning, incident locations, and outdoor recreation searches. One type of geospatial website is the online real estate search. Many realty websites allow prospective residential property buyers to sort listed properties interactively based on desired elements. These elements typically address features wanted within a home, such as the dwelling's size, number of bedrooms and bathrooms, whether a garage or swimming pool is included, and other furnishings. Equally important in considering the ideal home is to find the ideal location. Length of the commute time, crime frequency, proximity to cultural and retail options, and the location of desired schools can provide for an overall ""neighborliness"" that is vital to ensure a comfortable life in the new home at the new location. While some websites are beginning to address this concern by including small overlays within a property's webpage, none overtly considers that the home buying process may not start with the selection of home features, but by first determining the ""best-fit"" neighborhood. The web application created for this thesis is unique in its premise to first introduce potential homebuyers to neighborhoods. Prospective homebuyers may select from several neighborhood factors to find locations that satisfy their search parameters. An overlay of available properties is then displayed for the web application user to show what offerings are available in those resulting areas.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

GIS data curation and Web map application for La Brea Tar Pits fossil occurrences in Los Angeles, California

PDF

Wake County District Overlay: an online electoral data visualization application

PDF

Guiding business oriented volunteered geographic information through geotigger services: a case study of CrossFit affiliates

PDF

Social media canvassing using Twitter and Web GIS to aid in solving crime

PDF

Semi-automated visualization of spatial information in unstructured text

PDF

A user study of GIS infused genealogy with dynamic thematic representation and spatiotemporal control

PDF

Geological modeling in GIS for petroleum reservoir characterization and engineering: a 3D GIS-assisted geostatistics approach

PDF

Development of a mobile GIS high-water mark data collection application for the Mississippi River Basin

PDF

Evaluating the utility of a geographic information systems-based mobility model in search and rescue operations

PDF

Silicon Valley construction project web mapping application

PDF

Development of a Web GIS for urban sustainability indicators of Oakland, California

PDF

The spatial effect of AB 109 (Public Safety Realignment) on crime rates in San Diego County

PDF

Analyzing earthquake casualty risk at census block level: a case study in the Lexington Central Business District, Kentucky

PDF

Implementing spatial thinking with Web GIS in the non-profit sector: a case study of ArcGIS Online in the Pacific Symphony

PDF

Development of a Web GIS application to aid marathon runners in the race selection and planning process

PDF

Residential housing code violation prediction: a study in Victorville, CA using geographically weighted logistic regression

PDF

Integration of topographic and bathymetric digital elevation model using ArcGIS interpolation methods: a case study of the Klamath River Estuary

PDF

Selection of bridge location over the Merrimack River in southern New Hampshire: a comparison of site suitability assessments

PDF

A threat-based least-cost path decision support model for national security resource allocation along the US-Mexico border

PDF

Concentrated solar thermal facilities: A GIS approach for land planning

Asset Metadata

Creator Dowling, Jennifer Nicole (author)

Core Title Finding your best-fit neighborhood: a Web GIS application for online residential property searches for Anchorage, Alaska

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 09/23/2014

Defense Date 09/08/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag GIS,neighborhood,neighborhood search,OAI-PMH Harvest,online real estate,Real estate

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Swift, Jennifer N. (committee chair), Chiang, Yao-Yi (committee member), Lee, Su Jin (committee member)

Creator Email jndowlin@usc.edu,jndowling@gci.net

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-484539

Unique identifier UC11286719

Identifier etd-DowlingJen-2979.pdf (filename),usctheses-c3-484539 (legacy record id)

Legacy Identifier etd-DowlingJen-2979.pdf

Dmrecord 484539

Document Type Thesis

Format application/pdf (imt)

Rights Dowling, Jennifer Nicole

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