A paper recommender system based on user’s profile in big data scholarly

Abstract

Users encounter a huge volume of papers in digital libraries and paper search engines such as IEEE Explore, ACM Digital library, Google scholar and etc. these high number of papers make some difficulties for researchers for finding proper information and items. Recommender systems contain successful tools for knowledge of users and identification of their priorities. These systems present a personalized proposal to users who seek to find a special kind of relevant data or their priorities through the big number of data. Recommendersystem based on personalization uses the user profile and in view of the fact that the user profile encompass information pertaining to the user priorities; so it is a very active district in data recovery. Recommendersystem is an attitude that presented in order to encounter difficulties caused by abundant data and it helps users to attain their goals quickly through huge number of data. In this paper, we have presented an approach that received entire of available information in a paper, and formed a profile for each user by short and long inquiries; this profile is personalized for user and then recommends the closest paperto the  user by the user profile. Findings indicate that suggested approach outperformsthe similar approaches.

Keywords: recommender system; bigdata; user profile; content-based recommender system; hadoop