Hybrid Data Aggregation Technique to Categorize the Web Users to Discover Knowledge About the Web Users

• Published in: Wireless personal communications 2017-12, Vol.97 (4), p.5289-5303
• Main Authors: Manohar, E., Shalini Punithavathani, D.
• Format: Article
• Language: English
• Subjects: Advertisements; Algorithms; Communications Engineering; Computer Communication Networks; Customization; Data management; Data mining; Data transfer (computers); Engineering; Human behavior; Networks; Signal,Image and Speech Processing; Software reviews; Success; Traffic information; User satisfaction; Websites; World Wide Web
• ISSN: 0929-6212; 1572-834X
• DOI: 10.1007/s11277-017-4779-x

Web usage mining is a knowledge discovery technique where a data analyst can discover useful information from the web users’ data. Web contains billions of web pages. The web access behaviour of one web user differs from that of another and also it changes with respect to their temporal property. By analyzing the users’ data, the web administrator can personalize the web pages according to individual web users’ interest. Personalizing the web page gives various advantages in this fast era such as low search time, less data transfer, higher availability of data, lower bandwidth traffic, targeted advertisement and identifying the threaded web users and high web users satisfaction. Due to the above advantages, it is very much essential in the present World Wide Web. Various algorithms, techniques and tools are available in the field of web usage mining. Although there are various techniques, algorithms and tools developed related to web usage mining, new techniques are required to make the discovery of knowledge more accurate. In this paper, a novel technique is proposed by using various methods such as web log, web ranking, web rating and web review based method to identify the success rate of various web pages and summarize the value to identify the accurate success rate of each web page. The success rate is normalized and aggregated into three categories for personalizing the web user. Personalizing the web user based on grouping relevant web access behaviour reduces the calculation complexity. It is very effective in very large websites. This technique is very much effective for analyzing the outreach of web advertisement to the web users.