Recommender Systems to Address New User Cold-Start Problem with User Side Information
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2016, Vol 18, Issue 2
Abstract
Abstract: Due to exponential growth of Internet, users are facing the problem of Information overloading. Recommender Systems (RS) serve as an indispensible tool to solve information overloading problem. Due totheir great commercial value, recommender systems have also been successfully deployed in industry, such as product recommendation at Amazon, music recommendation at iTunes, movie recommendation at Netflix, etc. Collaborative Filtering (CF) is one of the popular approaches for Recommender systems based on user-item rating matrix. The main challenges faced by CF are Sparsity, Cold-Start and Scalability. In this paper we proposed a new approach to use user’s side information in addition to user-item rating matrix to address new user cold-start problem. User’s side information is obtained from Social Networks, which are the most activesource of information now a day. First CF method is used to form user clusters based on the similarity of users. User’s side information is obtained from Social networks and the same is used to form a Social matrix. Finallycombine user history with social matrix to provide recommendations to new users. The experimental results proved that the performance of RS is improved over traditional CF systems.
Authors and Affiliations
M. Sunitha , Dr. T. Adilakshmi
Keywords
Related Articles
- EP ID EP101355
- DOI -
- Views 96
- Downloads 0
Protection of Direct and Indirect Discrimination using Prevention Methods
Along with privacy, discrimination is a very important issue when considering the legal and ethical aspects of data mining. It is more than observable that the majority people do not want to be discriminated &nbs...
Using Data-Mining Technique for Census Analysis to Give GeoSpatial Distribution of Nigeria.
There are patterns buried within the mass of data in the various editions of population census figures in this country. These are patterns that will be impossible for humans working with bare eyes and hands, to u...
A Review of Optical Character Recognition System forRecognition of Printed Text
Abstract: As we move ahead in technology advancements, from simple data processing, to intelligentcomputing, one area of research undergoing advancement, is the system of reading text characters on an image.Optical...
Modified Method of Generating Randomized Latin Squares
In combinatorics and in experimental design, Latin square (LS) plays a vital role. It has been studied for centuries and especially used in statistics. Let L(n) be the number of Latin square of order n . L(n) incre...
Object Elimination and Reconstruction Using an Effective Inpainting Method
Three major problems have been found in the existing algorithms of image inpainting: Reconstruction of large regions, Preference of filling-in and Choice of best exemplars to synthesize the missing reg...