1Research and Development Centre, Bharathiyar University, Coimbatore
2Department of Information Technology, K. Ramakrishnan College of Engineering, Trichy
3Department of Computer Applications, PSNA College Of Engineering And Technology., Dindigul
4Department of Computer Science, Government Arts and Science College, Veppanthattai-621116, Perambalur-dt, Tamil Nadu, India
Corresponding author email: smk76dgl@gmail.com
Article Publishing History
Received: 11/10/2020
Accepted After Revision: 27/12/2020
Web usage Mining is a kind of the web analysis, pre-processing stage in WUM consists the following Data Cleaning, user identification, session identification and path completion (path added). In this research article focuses the path completion part. Web log is the most valuable input of the web analysis with WUM. Web log information is collected from the server, client and proxy server; in these missed some of the information’s. It results missing access references, user access patterns are not clearly identified by incomplete access log. To rectify these issues path completion takes the role to acquire the missing reference. Different approaches are available to find the missing reference like Url, IP address, reference length but no one can be efficient. But our proposed candidate key constrained clustering algorithm for path completion can utilize the both type web log file client and server log file and also provide the better results in efficiency rather than the other path completion and construction techniques.
Web Usage Mining, Pre-Processing, Path Completion, Candidate Key, Clustering.