Traditionally, automated triage of papers is performed using lexical (unigram, bigram, and sometimes trigram) features. This pa- per explores the use of information extrac- tion (IE) techniques to create richer linguistic features than traditional bag-of-words models. Our classifier includes lexico-syntactic pat- terns and more-complex features that repre- sent a pattern coupled with its extracted noun, represented both as a lexical term and as a semantic category. Our experimental results show that the IE-based features can improve performance over unigram and bigram fea- tures alone. We present intrinsic evaluation results of full-text document classification ex- periments to determine automatically whether a paper should be considered of interest to biologists at the Mouse Genome Inform...

Click Here to View the Paper In XML Format