The Personalized Product Recommendation System Based-On Users’ Reviews

With the rapid development of the Internet, there are more and more propducts on the site, and everyday the amout of the products is rising exponentially,which has led to "information overload",the user is difficult to quickly and accurately locate products of interest,the personalized recommendation systems came into being. It aimed to recommend products that user maybe interested in. As a carrier, the product has its own objective data, but also contains a large number of subject information. In order to improve user experience, this thesis design and implement a personalized· product recommendation system by using users’reviews.In this paper propose a content-based recommendation algorithm based-on user comments,by mining users’reviews about products,the establishment of the vector space model of the product documents, use Latent Dirichlet Allocation (LDA) reducing the dimension for the model,ease of calculation the similarity between the product and combining user rating weights,finally recommend products for the target user. We evaluate the recommendation method based on precision, recall, F-measure on ten movie recommendations. The experimental results show that the proposed algorithm can improve the recommendation accuracy comparing with the traditional content-based recommendation algorithm,and improve quality of recommendation greatly.