International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Mining the data from large-plain text enhances the retrieval of information from resources. This paper describes about the automatic discovery of part-whole patterns from the texts using knowledge. The Parts are found by learning semantic constraints and linking documents to the knowledge-base. Knowledge discovery in text is a potential method, which automatically extracts the concepts and conceptual relationships from text. The importance of part-whole relation in various application domains has been analyzed. In addition, performance can be improved by using Wikipedia as a knowledge-base, and the effect of semantic-drift is eliminated.