Bidirectional Search on Cyclic Behavioral Patterns for Web Sequential Mining

Web sequential pattern mining is significant for evaluating the access behavior of web users. Many research works have been developed for web sequential patterns mining. But, conventional web sequential patterns mining techniques takes more time for extracting frequent web pages. Besides, the performance of existing web sequential patterns mining was not efficient. In order to solve these limitations, Bidirectional Expansion Search based Cyclic Behavioral Pattern mining (BES-CBPM) method is proposed. The BES-CBPM method is designed with objective of increasing the mining efficiency of web sequential patterns with minimum time and space complexity. The BES-CBPM method used Bidirectional Expansion Search algorithm for extracting the similar users interest web pages from weblog database. The Expansion Search algorithm developed in BES-CBPM method simultaneously carried outs two search with aim of mining the similar users interest web pages. This process resulting in reduced time and space complexity for web sequential patterns mining. In addition, BES-CBPM method used cyclic behavioral search based pattern mining with aiming at enhancing the mining efficiency. The cyclic behavioral search based pattern mining in BES-CBPM method reiterates the bidirectional expansion search for a certain period of time in order to efficiently improve mining performance of web sequential patterns mining. The BES-CBPM method conducts the experimental works on metrics such as time complexity, space complexity, mining efficiency, true positive rate of mining and scalability. The experimental results demonstrate that the BES-CBPM method is able to improve the mining efficiency and also reduces the time complexity of web sequential pattern mining when compared to state of the art works.