• P-ISSN 0974-6846 E-ISSN 0974-5645

Indian Journal of Science and Technology

Article

Indian Journal of Science and Technology

Year: 2015, Volume: 8, Issue: 29, Pages: 1-9

Original Article

Scalable Recommendation Engine for Optimized Product Discovery

Abstract

Recommending products to the user in e-commerce sites must be based on the user’s taste and buying pattern. When a new user steps into the site, the system is unable to generate recommendations for that user, which is termed as cold start problem. The main objective is to introduce an approach that provides customized recommendations even to a new user thereby solving the cold start problem. Both the cold start problem and customized recommendation approach in the system can be addressed by combining two approaches. The first approach is building a graph relation which helps in knowing the taste of the user and thus overcomes the cold start problem and the second approach is harnessing the hidden potential from the review corpus of e-commerce sites. Review corpus consists of details such as rating, reviews of various users, each with a peculiar taste. When the user’s taste and opinion polarity of the product is converged together it will lead to an optimized product recommendation to the user. The system uses hadoop, a scalable framework which enables to get recommendation in near real time. Orient DB is used for building graph relations. During testing this approach worked well with the cell phone accessories and clothing domain.
Keywords: Content based filtering, Opinion Mining, Recommendation Engine, Sentimental Analysis

DON'T MISS OUT!

Subscribe now for latest articles and news.