The Fuzzy Believer System

Revision with unchanged content. The growing number of publicly available information sources makes it impossible for individuals to keep track of all the various opinions on one topic. The goal of our artificial believer system presented in this book is to extract and analyze statements of opinion from newspaper articles. Beliefs are modeled using a fuzzy-theoretic approach applied after natural language processing-based information extraction. A fuzzy believer models a human agent, deciding what statements to believe or reject based on different, configurable strategies. The book aims to provide an overview of the problems related to computational processing of natural language in the specific domain of newspaper articles and presents a system which offers one possible solution to deal with these problems. It is therefore suited for readers interested in Natural Language Processing and the implementation of an application in this area.