Sentential association based text classification systems

Abstract

We recently proposed a novel sentential association based approach SAT-MOD for text classification, which views a sentence rather than a document as an association transaction, and uses a novel heuristic called MODFIT to select the most significant itemsets for constructing a category classifier. Based on SAT-MOD, we have developed a prototype system called SAT-Class. In this demo, we demonstrate the effectiveness of our text classification system, and also the readability and refinability of acquired classification rules. © Springer-Verlag Berlin Heidelberg 2005.