Secure and Efficient Web Usage Mining : A Novel Framework for Prediction of Consumer Behaviour

Abstract

To protect the weblog data from various threats of active attacks security of data is the major challenge for the organization. To obtain maximum profit keeping intact the interest of consumers is some other major aspect of an organization. Web usage mining techniques prove a benchmark in the field of web development consumer behaviour web intelligence. In the past to analyze consumers behaviour conventional Apriori algorithm was used and improved version i.e. T_Apriori algorithm was introduced by the researcher but problems of space complexity time complexity and no concept of security of the data is found in these algorithms. The present paper presents a Novel Secure and Efficient Web Usage Mining(SEWUM)which not only ensure the security of data but also accurately predicts consumer behaviour with the security of data by exhausting the Hash table Hashing Encryption Message-Digest (MD5) algorithm encoding and decoding and anti-monotonicity Apriori property has a distinct edge over the classical Apriori algorithm and T_Apriori algorithm. The performance evaluation of the proposed system is carried out in python on real data (weblogs). Experimental results show that the proposed technique is more effective for the organization and results can be obtained in less memory consumption and less execution time.