AN ADVANCED PRODUCT RECOMMENDATION FRAMEWORK BY USING MULTIPLE PRODUCT FEATURE ANALYTICS

Online store plays a vital role for product analysis using automatic recommended system to the new customers. Since purchase patterns of traditional approaches considered the overall product rating and website preferences in the recommendation process. Commercial ecommerce recommended system use machine learning techniques to make appropriate decision rules for the customers during real-time sessions. Also, various machine learning techniques depend on the type of product and the number of transactions. It also depends on the customer purchase history whose purchase rules are close to that of new customers. Conventional recommendation techniques generate recommendation patterns similar to products that the target user has computed recommendation by analyzing the products purchased by the customers who are identical to the target customers. In this proposed work, an improved multi-feature based product recommendation system was built on the real time ecommerce sites. In this proposed architecture, multiple web based products are analyzed and ranked by using multi-product based recommender system. In this system, multiple products from different vendors are taken with multiple product features. The proposed work gives the best solution to the users who are interested in comparing the different vendors' products for product purchase. Experimental results show that proposed multi-product system give better user product recommendation compared to traditional single site product recommendation systems.