Formal distance vs. association strength in text processing

Abstract

Text information processing depends critically on the proper document representation. Traditional models, like vector space model, have significant limitations because they do not consider semantic relations amongst terms. In this paper we analyze a document representation that uses an association graph scheme model called Global Association Distance Model or GADM, the significance of the formal distance for the association strength, and the application of several distance-strength functions in this model. We evaluate this significance for topic classification tasks. © Springer-Verlag Berlin Heidelberg 2007.