DRec:Multidomain Recommendation System for Social Community

Abstract

The Recommendation System is the software engines and the approaches for consideration proposal to the user which might be most probably matched with the liking of users. Usually, Recommendations system recommends on various fields like what items to buy, which movies to watch even the job recommendations, depending upon the users profile. Instinctively if the domains of users are captured and filtered out accordingly to recommend them will be a very useful idea. In this paper we will be discussing about the research done by us and the limitation of the system.We design a system for recommending domains in social network, using an explicit / offline data. We have tested it on two popular dataset namely Epinions and Ciao. The ratings of items are studied and performance measures are calculated with three different ways 1) MAP (Mean Average Precision) 2)F-measure and 3) nDCG (Normalized Discounted Cumulative Gain) . As well we have compared the results with 5 comparisons methods.All the techniques and methods are explained in paper.
DOI: 10.17762/ijritcc2321-8169.150810