Machine learning approach for argument extraction of bio-molecular events

Abstract

The main goal of Biomedical Natural Language Processing (BioNLP) is to capture biomedical phenomena from textual data by extracting relevant entities and information or relations between biomedical entities such as proteins and genes. Previous research was focussed on extracting only binary relations, but in recent times the focus is shifted towards extracting more complex relations in the form of bio-molecular events that may include several entities or other relations. In this paper we propose a machine learning approach based on Conditional Random Field (CRF) to extract the arguments of bio-molecular events. The overall task involves identification of event triggers from texts, classification of them into some predefined categories and determining the arguments of these events. We identify and implement a set of features in the forms of statistical and linguistic features that represent various morphological, syntactic and contextual information. Experiments on the benchmark setup of BioNLP 2009 shared task show the recall, precision and F-measure values of 45.75%, 78.93% and 57.91%, respectively.