A New Preprocessing Approach for Mining Association Rule from Large Database
Kalpana Wani1, J. W. Bakal2
1Prof. Kalpana Wani, Department of Computer Engineering, Mumbai University/PIIT College, New Panvel,Navi (Maharashtra), India.
2Dr. J.W. Bakal, Department of IT, Mumbai University/ SS Jondhale College of Engineering, Dombivali, Mumbai (Maharashtra), India.
Manuscript received on 20 March 2014 | Revised Manuscript received on 25 March 2014 | Manuscript published on 30 March 2014 | PP: 55-60 | Volume-3 Issue-1, March 2014 | Retrieval Number: A0994033114/2014©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: As we know now a day’s internet plays vital role in serving the needs of the user’s on web. We can search the information, we can order the items and can do many things online. Different options are available for this, but many times we don’t have time to go through all alternatives and decide the best one. In such case a Web Page Recommendation System will be helpful to suggest the pages which are most relevant to your current search. The Server log files are generated as a result of an interaction between the client and the service provider on web. Server log file contains the massive hidden valuable information related to the visitors, if we mined this, it can be used for predicting the navigation behavior of the users. However the task of discovering frequent sequence patterns from the server log is challenging as it consist of huge data. Most of the time this data is incomplete and because of that it can’t be processed further for generating accurate knowledge. Proposed system focuses on adopting an intelligent technique that can provide personalized web service for accessing related web pages more efficiently and effectively. Proposed system uses two intelligent algorithms for predicting the user behavior’s namely FP Growth and Éclat. These algorithms save the time and space problem of existing system. Further from the frequent pages pattern Direct and Indirect Association Rules are generated and based on that Ranking is provided to pages which will help recommendation system to recommend similar search pages. This paper focuses on new approach for preprocess web log data.
Keywords: Association Rule, Indirect Association Rule, Parsing, Preprocessing.
Scope of the Article: Data Mining and Warehousing