ABSTRACT
ABSTRACT: This paper introduces Semantic Role Labeling (SRL) method for Semantic Identification (SI) and Conceptual Graph techniques (CG) for semantic representation. In Semantic Role Labeling, the semantic roles have been identified as agent, patient etc using the Karaka theory. The performance of SRL is calculated using Yet Another Chunk Annotator, a Support Vector Machine based algorithm which has shown significant improvement over earlier methods. The semantic representation introduced here is a directed graph which shows the relation between concepts and semantic roles.
