An Improved Intellectual Analysis Precedence and Storage for Business Intelligence from Web Uses Access Data

With the growth of data mining web usage, user behaviour analysis is a useful area for business intelligence. There are several techniques to extract interesting pattern and knowledge which will be used for business intelligence from user's access records. However, analysis of large Web log files is a convoluted task not fully addressed by existing web access techniques. In order to provide better storage and user behaviour from huge datasets the proposed intellect storage, precedence analysis (ISPA) algorithm has been introduced. The user session and history are considered as feedback for clustering. The proposed system considers the total number of hits, time spent by the user on a particular page and links. Based on these parameters, personalization has been proposed. The implementation of an effective pruning technique and FP-growth algorithm has provided better results and performance. This also considers outlier detection in order to group the links effectively. Experimental results are presented using user click through logs to validate the effectiveness of the proposed methods.