Semi-Supervised Classification for Extracting Protein Interaction Sentences using Dependency Parsing

Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pp. 228–237, Prague, June 2007. c⃝2007 Association for Computational Linguistics Semi-Supervised Classification for Extracting Protein Interaction Sentences using Dependency Parsing G¨unes¸ Erkan University of Michigan gerkan@umich.edu Arzucan ¨Ozg¨ur University of Michigan ozgur@umich.edu Dragomir R. Radev University of Michigan radev@umich.edu Abstract We introduce a relation extraction method to identify the sentences in biomedical text that indicate an interaction among the protein names mentioned. Our approach is based on the analysis of the paths between two protein names in the dependency parse trees of the sentences. Given two dependency trees, we define two separate similarity functions (ker- nels) based on cosine similarity and edit dis- tance among the paths between the protein names. Using these similarity functions, we investigate the performances of two classes of learning algorithms, Support Vector Ma- chines and k-nearest-neighbor, and the semi- supervised counterparts of these algorithms, transductive SVMs and harmonic functions, respectively. Significant improvement over the previous results in the literature is re- ported as well as a new benchmark dataset is introduced. Semi-supervised algorithms perform better than their supervised ver- sion by a wide margin especially when the amount of labeled data is limited. 1 Introduction Protein-protein interactions play an important role in vital biological processes such as metabolic and signaling pathways, cell cycle control, and DNA replication and transcription (Phizicky and Fields, 1995). A number of (mostly manually curated) databases such as MINT (Zanzoni et al., 2002), BIND (Bader et al., 2003), and SwissProt (Bairoch and Apweiler, 2000) have been created to store pro- tein interaction information in structured and stan- dard formats. However, the amount of biomedical literature regarding protein interactions is increas- ing rapidly and it is difficult for interaction database curators to detect and curate protein interaction in- formation manually. Thus, most of the protein in- teraction information remains hidden in the text of the papers in the biomedical literature. Therefore, the development of information extraction and text mining techniques for automatic extraction of pro- tein interaction information from free texts has be- come an important research area. In this paper, we introduce an information extrac- tion approach to identify sentences in text that in- dicate an interaction relation between two proteins. Our method is different than most of the previous studies (see Section 2) on this problem in two as- pects: First, we generate the dependency parses of the sentences that we analyze, making use of the dependency relationships among the words. This enables us to make more syntax-aware inferences about the roles of the proteins in a sentence com- pared to the classical pattern-matching information extraction methods. Second, we investigate semi- supervised machine learning methods on top of the dependency features we generate. Although there have been a number of learning-based studies in this domain, our methods are the first semi-supervised efforts to our knowledge. The high cost of label- ing free text for this problem makes semi-supervised methods particularly valuable. We …

Content truncated to 500 words