Due to the rapid growth of information available on the WWW (World Wide Web) and the rapid introduction of new web services, information overload becomes a critical problem that users are overwhelmed, and how to help users find the information they are interested in becomes an unprecedented challenge, attracting the attention of both research and application areas. In this paper we propose a new recommendation technique for the trust aware recommendation in social networks based on the Deep Learning (DL). Here, we firstly find out the per-train initial value of the parameter for this we used a deep auto-encoder. The community detection algorithm based on trust relations in social networks is proposed for the revamp the MF (Matrix Factorization) model with social trust together and community effect. This system deals with the cold-start users far more effeciently.
Pooja Shinde has completed her Masters in Engineering (Computer Engineering) from Savitribai Phule Pune University in Maharashtra, India. She has over four years of experience in the field of education and is currently working as a Lecturer in Dr. D Y Patil School of Engineering, Pune. She has guided several projects in various domains till date.
Number of Pages:
LAP LAMBERT Academic Publishing
Social Network, Recommendation Techniques, Deep Learning, Cold Start Problem, matrix factorization
TECHNOLOGY / General