Review Based Rating Prediction

Recommendation systems are an important units in today’s e-commerce applications, such as targeted advertising, personalized marketing and information retrieval. In recent years, the importance of contextual information has motivated generation of personalized recommendations according to the available contextual information of users. Compared to the traditional systems which mainly utilize users’ rating history, review-based recommendation hopefully provide more relevant results to users. We introduce a review-based recommendation approach that obtains contextual information by mining user reviews. The proposed approach relate to features obtained by analyzing textual reviews using methods developed in Natural Language Processing (NLP) and information retrieval discipline to compute a utility function over a given item. An item utility is a measure that shows how much it is preferred according to user’s current context. In our system, the context inference is modeled as similarity between the users reviews history and the item reviews history. As an example application, we used our method to mine contextual data from customers’ reviews of movies and use it to produce review-based rating prediction. The predicted ratings can generate recommendations that are item-based and should appear at the recommended items list in the product page. Our evaluations suggest that our system can help produce better prediction rating scores in comparison to the standard prediction methods. Introduction In recent years, recommendation systems (RecSys) have been extensively used in various domains to recommend items of interest to users based on their profiles. RecSys are an integral part of many online stores such as Alibaba.com, Amazon.com, etc. One of the most famous examples of a recommendation system is Amazon [4]. This system contains movie ratings for over 100,000 movies. A user’s profile is a reflection of the user’s previous selections and preferences that can be captured as rating scores or textual review given to different items in the system. Using preference data, different systems have been developed to produce personalized recommendations based on collaborative filtering, content-based or a hybrid approach. Despite the broad used of such recommendation systems, they fail to consider the users’ latent preferences, thus may result in performance degradation. For example, a customer who has once viewed a movie with his friend’s child may repeatedly receive suggestions to view kid’s movies as the recommendation algorithm select base on the whole history in user’s profile without prioritizing his interests. To address this issue, review-based recommendation systems has been introduced. Contextual information about a user preference can be explicit or implicit and can be inferred in different ways such as user score ratings or textual reviews. We concentrate on deriving context from textual reviews. As an example application of our approach, we have used our method to mine contextual data from customers’ reviews of movies domain. In order to evaluate our method, we have used Amazon movies reviews [4]. The reason for choosing this dataset is that users usually provide some contextual cues in their comments. For example, they may mention that they are very fond of a specific actor or director, or they may express their opinions about the 1 ar X iv :1 60 7. 00 02 4v 3 [ cs .I R ] 2 7 Ju l 2 01 6