Detecting the evolution of semantics and individual beliefs through statistical analysis of language use

Reference:

Bilovich, A. and Bryson, J. J., 2008. Detecting the evolution of semantics and individual beliefs through statistical analysis of language use. In: Beal, J., Bello, P., Cassimatis, N., Coen, M. and Winston, P., eds. Naturally-Inspired Artificial Intelligence - Papers from the AAAI Fall Symposium, Technical Report. Arlington, VA: AAAI Press, pp. 21-26.

Related documents:

Official URL:

http://www.aaai.org/Papers/Symposia/Fall/2008/FS-08-06/FS08-06-004.pdf

Abstract

Individual differences in semantics and beliefs have up to now been identified primarily by questioning people. However, semantics and beliefs can also be observed in concrete, quantifiable contexts such as reaction-time experiments. Here we demonstrate an automatic mechanism which can replicate such semantics by observing regularities in language use through statistical text analysis. We postulate that human children, who are fantastic pattern recognizers, may also exploit this same information, thus our mechanism may be an essential module in a human-like cognitive system. In this article we first review the underlying theories and existing results, then present the tool itself. We validate the tool against existing semantic priming reaction-time results. Finally we use the tool to explore the evolution of beliefs extracted from three sources: the Bible, the works of Shakespeare and the contemporary British National Corpus.

Export

Actions (login required)