AUTOMATED PRODUCT RECOMMENDATION

AUTOMATED PRODUCT RECOMMENDATION

¹GANESHA C, ²PRAVEEN K S

¹Student, Department of Master computer Application, East West Institute of Technology, Bangalore, Karnataka, India

²Assoc. Professor, Department of Master computer Application, East West Institute of Technology, Bangalore, Karnataka, India

------------------------------------------------------***------------------------------------------------------

ABSTRACT

Any modern online retail or social networking site has a recommendation mechanism. As an example of outdated recommendation systems, the mechanism for product recommendations has two key flaws: suggestion duplication and unpredictability when it comes to new things (cold start). These limitations arise because traditional recommendation systems depend solely on the user's prior purchasing activity to provide new item recommendations. Meta-Interest is a personality-based a technique for recommending products depending on metapath discovery and user interest mining. The suggested strategy can improve the recommendation system's accuracy and recall, particularly in cold-start situations circumstances, according to experimental data.