Automatic extraction of cause effect relations from tamil Discourse

Abstract

The objective of the present work is to study the cause effect newlinerelation in Tamil Discourse develop a computational grammar and use it to newlineautomatically extract cause effect relations in any Tamil text newlineA cause effect relation is a relation established between two newlinepropositions in a language concerning two events or situations in such a way newlinethat the event described by one of the propositions is communicated as newlinebringing about the event proposed by the other proposition In other words newlinethe first proposition is the cause and the second proposition is the effect newlineRecognizing and extracting such cause effect relations from texts would aid newlinein the development of natural language processing applications like questionanswering newlinesystems and reasoning systems newlineIn this work a corpus in Tamil was identified and a corpus study newlinewas done to find the cause effect markers that function at different levels of newlinelinguistic structure The cause effect markers were classified hierarchically newlinebased on their linguistic features and using this classification a computational newlineclassification was arrived at Hence a computational grammar for cause effect newlinerelations was defined The grammar was tested by using it to extract the newlinefeatures for machine learning newline newline