Rules based event extraction from natural language text

Abstract

The notion of events plays a crucial role in narrative texts. Events are the situations that happen or occur at a particular place and time. Extraction and representation of event has a significant role in many of the natural language text and applications like text summarization, question answering systems etc. Several methods were developed so far but they addressed the problem in domain aspect point of view. In this paper we demonstrate a rule based system for event extraction which is generic in nature. A generic rules based event extraction algorithm was developed that maps the framed rules to derive the events from natural language text. Experiments were conducted on the MUC data set and the results are found to be encouraging.