Collaborative Filtering with Preventing Fake Ratings

Social voting is a promising new feature in online social networks. It has distinctive challenges and opportunities for suggestion. In this paper, we increase a set of matrix factorization (MF) and nearest-neighbor (NN)-based recommended systems (RSs) that explore user social network and group association information for social voting recommendation. During experiments with actual social voting traces, we express that social network and group association information can drastically progress the popularity-based voting advice, and social network in order dominates group association sequence in NN-based approaches. We as well observe that social and group information is much more precious to cold users than to heavy users. In our experiments, simple meta path based nearest-neighbor models outperform computation-concentrated on matrix factorization models in hot-voting recommendation, while user's preferences for non-hot votings can be better mined by matrix factorization models. We further put forward a hybrid RS, bagging distinct single approaches to get the best top-k hit rate. Dr. A. Srinivasa Rao | B. Bhagyalakshmi | Ab. Sirajunnisa | Md. Ashraf | E. Harika | Ch. Gangadhar"Collaborative Filtering with Preventing Fake Ratings" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-2 | Issue-3 , April 2018, URL: http://www.ijtsrd.com/papers/ijtsrd11334.pdf http://www.ijtsrd.com/engineering/computer-engineering/11334/collaborative-filtering-with-preventing-fake-ratings/dr-a-srinivasa-rao