The technique of adapting a website to match the specific requirements of an individual user or a group of users is known as web personalization. Association rule mining and sequential pattern finding, as well as grouping and classification, are some of the ways it uses to make accurate predictions about users' future moves. In terms of precisely forecasting user interests, recent attempts to online personalization have fallen short. To provide meaningful ideas, a new online customising approach has been developed. The core of the suggested work is to give appropriate weights to the web pages of a website. We used the distance measure of the visit relationship as well as the occurrence frequency measure of web pages to accomplish this. For grouping web pages and identifying current user behaviours more precisely than earlier approaches, the strategy employs an improved graph-based partitioning algorithm. A Threshold value is used when deciding between web pages for the purpose of recommendation. The results of our tests suggest that accuracy is around 61%, coverage is around 34%, and the F1 measure is 44.23 percent. It contributes to the user's improved browsing experience.
Author(S) Details
N. Krishnaiah
Department of IT, St. Martin’s Engineering College, Secunderabad, Telangana, India.
R. Nagaraju
Department of IT, St. Martin’s Engineering College, Secunderabad, Telangana, India.
D. Babu Rao
Department of IT, St. Martin’s Engineering College, Secunderabad, Telangana, India.
B. Venkateswara Rao
Department of IT, B V Raju Institute of Technology, Narsapur, Telangana, India.
View Book:- https://stm.bookpi.org/NPER-V6/article/view/5499
No comments:
Post a Comment