Page 1 |
Save page Remove page | Previous | 1 of 209 | Next |
|
small (250x250 max)
medium (500x500 max)
large ( > 500x500)
Full Resolution
All (PDF)
|
This page
All
Subset |
CONTEXT DEPENDENT UTILITY:
AN APPRAISAL-BASED APPROACH TO MODELING CONTEXT, FRAMING,
AND DECISIONS
by
Jonathan Ito
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMPUTER SCIENCE)
May 2013
Copyright 2013 Jonathan Ito
Object Description
| Title | Context dependent utility: an appraisal-based approach to modeling context, framing, and decisions |
| Author | Ito, Jonathan |
| Author email | ito@ict.usc.edu;jonathan.ito@gmail.com |
| Degree | Doctor of Philosophy |
| Document type | Dissertation |
| Degree program | Computer Science |
| School | Viterbi School of Engineering |
| Date defended/completed | 2012-12-13 |
| Date submitted | 2013-02-01 |
| Date approved | 2013-02-01 |
| Restricted until | 2013-02-01 |
| Date published | 2013-02-01 |
| Advisor (committee chair) | Marsella, Stacy |
| Advisor (committee member) |
John, Richard Tambe, Milind Gratch, Jonathan |
| Abstract | Computational models of human decision behavior are becoming increasingly important for a wide range of applications. One challenge in creating these models is accounting for the effect of context on decision behavior. Research has shown that seemingly minor changes in the presentation, or framing, of a decision problem can lead to drastic shifts in behavior. ❧ Therefore, this work presents Context Dependent Utility (CDU), a comprehensive and novel framework for computationally modeling contextually-sensitive human decision behavior. CDU moves beyond existing approaches by: a) accurately accounting for a wider range of framing effects by modeling the multidimensional impact of context on decision behavior; b) providing principled and concrete guidance based on Appraisal Theory for identifying, evaluating, and encoding the salient aspects of context into the decision process; c) enabling the generalization of preferences across contextually distinct situations by exploiting the relationship between context and preferences; d) encouraging a high degree of interoperability in computational domains relying on utility-based models of human decision behavior by integrating appraisal with decision-theoretic techniques. ❧ Furthermore, this work presents empirical data supporting the behavioral predictions of CDU regarding the effect of context on decision behavior in two distinct decision scenarios involving school dropout prevention and job layoff prevention. The hypotheses are as follows: a) when outcomes are described as gains, outcomes will tend to be perceived as pleasant and decision behavior will tend towards risk-aversion when compared to outcomes described as losses; b) when high standards for success are established, outcomes will tend to be perceived as goal-incongruent and decision behavior will tend towards risk-seeking when compared to low standards of success; c) when the source of outcome uncertainty is depicted as arising from chance events, outcomes will tend to be perceived as uncontrollable and decision behavior will tend towards risk-aversion when compared to depicting the source of uncertainty as arising from one's ability. In addition to exploring the behavioral predictions of CDU, empirical data is employed in a comparative fitting analysis between several computational decision models commonly used to handle framing effects. The computational modeling results suggest that models which adopt a multidimensional view of context, such as CDU, can be much more effective at modeling decision behavior across distinct contextual environments than existing, one-dimensional models. In sum, the two empirical studies and the accompanying computational modeling analysis support the general hypothesis that different aspects of context can affect decision behavior in different ways. ❧ Finally, a real-world application of CDU in modeling a security-based decision scenario is discussed. In particular, the decision scenario is based on a Stackelberg Security game, representing a competitive scenario involving an attacker and a defender. The example illustrates how integrating considerations of context, through applying CDU, may allow for a more accurate model of the attacker decision behavior and ultimately empower the defender to make better decisions. Furthermore, the example also shows how CDU may provide a framework for strategically influencing an opponent through the manipulation of context. |
| Keyword | context; framing; utility; appraisal; emotion; decision; cognitive modeling; affective computing |
| Language | English |
| Part of collection | University of Southern California dissertations and theses |
| Publisher (of the original version) | University of Southern California |
| Place of publication (of the original version) | Los Angeles, California |
| Publisher (of the digital version) | University of Southern California. Libraries |
| Provenance | Electronically uploaded by the author |
| Type | texts |
| Legacy record ID | usctheses-m |
| Rights | Ito, Jonathan |
| 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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given. |
| Repository name | University of Southern California Digital Library |
| Repository address | USC Digital Library, University of Southern California, University Park Campus MC 7002, 106 University Village, Los Angeles, California 90089-7002, USA |
| Repository email | cisadmin@usc.edu |
| Archival file | uscthesesreloadpub_Volume6/etd-ItoJonatha-1423.pdf |
Description
| Title | Page 1 |
| Full text | CONTEXT DEPENDENT UTILITY: AN APPRAISAL-BASED APPROACH TO MODELING CONTEXT, FRAMING, AND DECISIONS by Jonathan Ito A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (COMPUTER SCIENCE) May 2013 Copyright 2013 Jonathan Ito |
Comments
Post a Comment for Page 1
