Kernel methods, syntax and semantics for relational text categorization

Abstract

Previous work on Natural Language Processing for Information Retrieval has shown the inadequateness of semantic and syntactic structures for both document retrieval and categorization. The main reason is the high reliability and effectiveness of language models, which are sufficient to accurately solve such retrieval tasks. However, when the latter involve the computation of relational semantics between text fragments simple statistical models may result ineffective. In this paper, we show that syntactic and semantic structures can be used to greatly improve complex categorization tasks such as determining if an answer correctly responds to a question. Given the high complexity of representing semantic/syntactic structures in learning algorithms, we applied kernel methods along with Support Vector Machines to better exploit the needed relational information. Our experiments on answer classification on Web and TREC data show that our models greatly improve on bag-of-words.