WEB MINING USING K-MEANS CLUSTERING AND LATEST SUBSTRING ASSOCIATION RULE FOR E-COMMERCE

User latency plays a significant role in e-commerce. This latency can be minimized by a priori predicting and fetching probable web pages for web users to run the e-commerce activities. Those prediction techniques are normally supported by clustering, classification and some association rules based on the data set of web logs of navigations, searching and attached web links with the e-commerce web pages. This paper proposes an integrated web page prediction technique by analyzing web users' previous navigational behavior. K-means clustering and latest substring association rule are considered for developing the proposed method of e-commerce web page prediction. The proposed method is evaluated by analyzing the precisions values of the output clusters using the proposed prediction technique.