Extraction of conceptual graphs using machine learning for the semantic representation of text

Abstract

The objective of our work is to develop a methodology to extract conceptual structures in a text document such that the semantics of the content is captured Conceptual graph formalism Sowa 1984 is used for this purpose in this work We have modified and extended the definition of concepts given by Sowa In this work what constitutes a concept from the semantic and computational perspective has been explored Sowa 1984 defines concepts in terms of percepts He states The process of perception generates a structure u called a CG in response to some external entity or scene e For every percept p there is a concept c called the interpretation of p He considers percepts as the basic unit for concept formation His work is more focussed on the structural part of graph formalism In our work what a concept is and how concepts are formed is analysed in detail We term the phrases or collocation of words which have unique meaning in a particular context as a Semantic Unit SU Semantic unit is considered as the basis for our concept definition We define For every semantic unit SU there is a concept c which is directly related to the semantics or meaning of SU newline