Tuning support vector machines for biomedical named entity recognition

Tuning Support Vector Machines for Biomedical Named Entity Recognition Jun’ichi Kazama† Takaki Makino‡ Yoshihiro Ohta∗ Jun’ichi Tsujii† ¶ † Department of Computer Science, Graduate School of Information Science and Technology, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan ‡ Department of Complexity Science and Engineering, Graduate School of Frontier Sciences, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan ∗ Central Research Laboratory, Hitachi, Ltd., Kokubunji, Tokyo 185-8601, Japan ¶ CREST, JST (Japan Science and Technology Corporation) Abstract We explore the use of Support Vector Ma- chines (SVMs) for biomedical named en- tity recognition. To make the SVM train- ing with the available largest corpus – the GENIA corpus – tractable, we propose to split the non-entity class into sub-classes, using part-of-speech information. In ad- dition, we explore new features such as word cache and the states of an HMM trained by unsupervised learning. Experi- ments on the GENIA corpus show that our class splitting technique not only enables the training with the GENIA corpus but also improves the accuracy. The proposed new features also contribute to improve the accuracy. We compare our SVM- based recognition system with a system using Maximum Entropy tagging method. 1 Introduction Application of natural language processing (NLP) is now a key research topic in bioinformatics. Since it is practically impossible for a researcher to grasp all of the huge amount of knowledge provided in the form of natural language, e.g., journal papers, there is a strong demand for biomedical information extraction (IE), which extracts knowledge automati- cally from biomedical papers using NLP techniques (Ohta et al., 1997; Proux et al., 2000; Yakushiji et al., 2001). The process called named entity recognition, which finds entities that fill the information slots, e.g., proteins, DNAs, RNAs, cells etc., in the biomedical context, is an important building block in such biomedical IE systems. Conceptually, named entity recognition consists of two tasks: identifica- tion, which finds the region of a named entity in a text, and classification, which determines the se- mantic class of that named entity. The following il- lustrates biomedical named entity recognition. “Thus, CIITAPROTEIN not only acti- vates the expression of class II genesDNA but recruits another B cell-specific coactivator to increase transcriptional activity of class II promotersDNA in B cellsCELLTYPE.” Machine learning approach has been applied to biomedical named entity recognition (Nobata et al., 1999; Collier et al., 2000; Yamada et al., 2000; Shimpuku, 2002). However, no work has achieved sufficient recognition accuracy. One reason is the lack of annotated corpora for training as is often the case of a new domain. Nobata et al. (1999) and Collier et al. (2000) trained their model with only 100 annotated paper abstracts from the MEDLINE database (National Library of Medicine, 1999), and Yamada et al. (2000) used only 77 annotated paper abstracts. In addition, it is difficult to compare the techniques used in each study because they used a closed and different corpus. To overcome such a situation, the GENIA cor- pus (Ohta et al., 2002) has been developed, and …

Content truncated to 500 words