Title: Multi-kernel learning-based recommender system using adaptive neuro-fuzzy inference system
Authors: A. Salman Ayaz; A. Jaya; Zameer Gulzar
Addresses: Department of Computer Applications, BS Abdur Rahman Crescent Institute of Science and Technology, Tamil Nadu-Chennai, India ' Department of Computer Applications, BS Abdur Rahman Crescent Institute of Science and Technology, Tamil Nadu-Chennai, India ' Department of Computer Applications, BS Abdur Rahman Crescent Institute of Science and Technology, Tamil Nadu-Chennai, India
Abstract: Nowadays, the personalised recommendations for a user is mainly build based on users rating. The drawback of online websites is that it presents many choices that result in consuming more time. Also it becomes strenuous for the user to find the right information or product from numerous search results. In this work, a new multi-kernel learning (MKL) is developed for providing the relevant information based on current user behaviour through his/her click stream data without explicitly asking for these data. Adaptive neuro-fuzzy inference system (ANFIS) approach is used for detecting the victor's stream of data in online sites. This method recommends a browsing option and matches the data to a specific user group that meet the requirements of a particular user at the required time. The experimental results reveal that the proposed MKL-ANFIS recommendation system (RS) improved the accuracy and coverage values approximately 3-4% compared to the traditional methods such as fuzzy C-means (FCM), enhanced graph-based partitioning (EGBP) algorithm and C-means and centre of gravity (CM & COG).
Keywords: adaptive neuro-fuzzy inference system; ANFIS; data mining; multi-kernel learning; MKL; recommender system.
International Journal of Intelligent Enterprise, 2021 Vol.8 No.4, pp.424 - 435
Received: 08 Feb 2020
Accepted: 01 Jul 2020
Published online: 06 Oct 2021 *