1. How to submit my research paper? What’s the process of publication of my paper?
The journal receives submitted manuscripts via email only. Please submit your research paper in .doc or .pdf format to the submission email: jiii@ejournal.net.
2. Can I submit an abstract?
The journal publishes full research papers...[Read More]

An Integrative Social Network and Review Content Based Recommender System

Haojie Ma and Dongsheng Che
Department of Computer Science, East Stroudsburg University, East Stroudsburg, PA 18301, USA

Abstract—Traditional collaborative filtering (CF) recommender systems (RS) suffer the problems of poor rating accuracy, cold start and data sparsity. Friendship information is not used in current recommender systems, but such information may be useful for improving rating accuracy. Besides, traditional recommender systems only use rating scores to predict users’ future buying decisions. Review comments are usually ignored, and thus it will decrease the recommendation accuracy. In this paper, a novel algorithm which integrates the social network (i.e. friendship information) and latent semantic analysis (i.e. review text) is proposed and implemented in our recommender system, aiming to improve rating accuracy. To test the performance of our recommender system, we crawled the restaurant data in the greater New York area from the website of Yelp. Our experimental results have shown that our proposed algorithm performs 18.6% higher accuracy than the traditional CF algorithm, and 1.2% higher accuracy than friend-based algorithm. We have also shown that the use of distant friendship information in the recommender systems could dramatically increase coverage rate.

Index Terms—big data, collaborative filtering, machine learning, review analysis, recommender system, social network

Cite: Haojie Ma and Dongsheng Che, "An Integrative Social Network and Review Content Based Recommender System," Journal of Industrial and Intelligent Information, Vol. 4, No. 1, pp. 69-75, January 2016. doi: 10.12720/jiii.4.1.69-75
Copyright © 2012-2018 Journal of Industrial and Intelligent Information, All Rights Reserved