Find results with:

Add another field

Search by date

to

Searching collections:

Dartmouth Dissertations

Add or remove collections

Home Dartmouth Dissertations Page 1

Reference URL

To link to the entire object, paste this link in email, IM or document

To embed the entire object, paste this HTML in website

To link to this page, paste this link in email, IM or document

To embed this page, paste this HTML in website

A framework of computational opinions / by Fei Yu.

you wish to report:

Your comment:

Your Name:

Submit

Cancel

...

Back to top

Powered by CONTENTdm®

Author	Yu, Fei
Title	A framework of computational opinions / by Fei Yu.
Date	2013
Abstract	We have a limited understanding of how an opinion is originated, how an opinion and information supporting and explaining it gets conveyed, and how the communicated opinion is perceived and processed by others. One direction of current research focuses on the conditions and determinants for opinion formation, while another focuses on opinion change induced by external influence (social influence in particualr), however, empirical findings have concluded that the persistence of opinion change is inconsistent. This indicates that current methods used to predict the next opinion based on the current opinion may fail if the decay in opinion change is rapid. We realize that for the existing computational framework/models, the close interplay between opinion formation and change is not exploited well. Prevelent computational frameworks/methods either model only on opinion formation process or only one opinion change process. Furthermore, computational frameworks/models that actually focus on opinion dynamics either only model one task or model a sequence of tasks but do not differentiate between them. Our insight to better address this problem is that we recognize the learning nature of opinion change and the decision-making nature of opinion formation: what has been learned through internalizing an external influence guides how decisions are made to externalizing cognitive processes. Thus the challenges for building a computational framework lies in modeling both the learning and decision-making aspects for the entire opinion formation task. To address this modeling challenge, we also recognize that the individual performing an opinion formation task is essentially engaged in a sequential decision-making problem with a specific goal in mind. This thesis presents our effort in two phases: In Phase I, we propose a networked space of reasoning processes which we call the Double Transition Model (DTM). Each mode within a DTM represents a cognitive state based on different degrees of query and knowledge incompleteness. The edges within a DTM denotes how query and knowledge differ between the connecting states which can be fulfilled to trigger changes in an opinion. To ensure that diversification in opinion formation processes can be covered by a DTM, we evaluated it by simulating four commonly accepted heuristics in human reasoning. The edges in a DTM are designed for modeling influence. In Phase I, we do not differentiate the causes of the influence - whether it is internally embraced or externally activated, so a DTM is mainly a cognitive model. In Phase II, we extend a DTM into a knowledge-based behavior model. How individuals perceive messages, accept messages, and communicate messages is simply a sequence of decisions. Therefore, we can construct a computational framework of opinions in which an opinion formation task can be formulated as a Markov decision process defined using an augmented DTM. Furthermore, we are able to use methods such as Q-learning to free us from the requirements of having full knowledge of communication dynamics in deriving optimal policies. In a case study, we demonstrate the power of the framework in making the entire opinion dynamics between two individuals analyzable.
Description	xiii, 203 leaves
Format	application/pdf
Advisor	Eugene Santos, Jr., George Cybenko, Mark Borsuk, Joseph M. Rosen, Hien Nguyen
Issuing Body	Dartmouth Dissertations are made available through the Dartmouth Digital Library.
Degree	Thesis (Ph.D.)--Dartmouth College.
Copyright	All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means without permission in writing from the author.
DOI	http://dx.doi.org/10.1349/ddlp.1016

	All
	Pages Enter page numbers or page ranges separated by commas. For example: 1,3,5-12

Find results with:

Searching collections:

Home Dartmouth Dissertations Page 1

A framework of computational opinions / by Fei Yu.

Page 1

Object Description

Description

A framework of computational opinions / by Fei Yu.

A framework of computational opinions / by Fei Yu.