Design and Implementation of Association Rules Based System for Evaluating WSD

Samit Kumar, Dr. S. Niranjan

Abstract: In this paper we present new method that used association rules mining techniques in the field of natural language processing. Word sense disambiguation is persistently a central and challenging problem as increasing usage of internet in daily life. Every user has some queries that have to be searched on the internet. Transactional Database is created after preprocessing the text files. Ambiguous words along with its context and Senses are stored in the SQL database to create transactional database. Apriori algorithm applied on this transactional database to mine the rules. Strong rules are generated using frequent itemsets. Sense of ambiguous word is deduced by using the strong association rules generated.

Keywords: Association rules, Apriori, Data mining, WordNet, Fuzzy Rules