Knowledge Discovery from Semi-Structured Data for Conceptual Organization

Conceptual organization of semi-structured documents can help in effective retrieval from collections of emails, product complaints, video descriptions etc. In this paper, we propose a conceptual organization scheme for grouping and categorizing semi-structured text data using natural language processing techniques. We propose a knowledge-discovery mechanism that extracts noun phrases from documents and arranges them into concept maps based on their co-occurrence. The emerging concept maps can be used for automatic grouping and conceptual categorization of documents. Further, Phrase structure Grammar is employed to extract relationships among these entities from documents and index the document collection with these relations.