Gerard, Christine (2000) Modelling readers of news articles using nested beliefs. Masters thesis, Concordia University.
Due to the wealth of documents available on-line, information retrieval on the Internet, in document databases or newspaper web sites can often lead to hundreds of documents being selected of which only a small number are of any interest to the user. This thesis is concerned with creating models of readers of news articles which could be used to filter and evaluate information to present to the reader only those articles containing information relevant to the reader's search. News articles are of special interest since they consist of information told to a reporter by game sources. Such reported speech can only be adequately modelled using nested belief models to represent the reader's beliefs about the reporter's beliefs about the source's beliefs. This thesis proposes a method for representing a news article, analysing it to extract encoded information, determining the reliability of the information reported in the article, creating models of all the agents in the article, and simulating how a reader of news articles acquires or adopts information from the sources in the article. A system called Percolator is presented. Percolator is a stand-alone implementation of one component of the more complex system required to model the nested beliefs of readers of news articles. It demonstrates how a technique called modified belief percolation can be used to simulate how a reader can acquire beliefs from the sources in an article.
|Divisions:||Concordia University > Faculty of Engineering and Computer Science > Computer Science and Software Engineering|
|Item Type:||Thesis (Masters)|
|Pagination:||x, 138 leaves : ill. ; 29 cm.|
|Degree Name:||Theses (M.Comp.Sc.)|
|Program:||Computer Science and Software Engineering|
|Thesis Supervisor(s):||Bergler, Sabine|
|Deposited By:||Concordia University Libraries|
|Deposited On:||27 Aug 2009 17:16|
|Last Modified:||08 Dec 2010 15:18|
Repository Staff Only: item control page
Downloads per month over past year