Finding Contextual Term from Index of Documents using Apriori Algorithm

Abstract

Data mining is a multidisciplinary field that has mainly introduced to process large database and discover useful information that could help in decision making. World Wide Web is a vast resource of interlinked hypertext documents which are accessed via the Internet. Web mining, an extension of data mining, employs techniques of data mining to documents on the internet. Association rule mining is a major technique in the area of data mining. Association rule mining finds frequent itemsets from a set of transactional databases. Apriori algorithm is one of the earliest algorithm of association rule mining. Apriori employs an iterative approach known as levelwise search.. The Apriori Algorithm for mining frequent itemsets for boolean association rules can be applied to the index file of a search engine to find contextually related terms i.e the terms which occur in number of documents together. Ranking, a major component of a search engine ranks the web pages based on some criteria. The ranking algorithms consider the keywords entered in the query for rank purpose. Considering contextually related terms with query term can improve the ranking system. In this paper it is shown how apriori algorithm can be applied on index of web documents to find contextually related terms.