Query Services Ranking for Recommendation Based on Review Analysis
Peng Li · Journal of Chinese Computer Systems · 2011
How to provide a personalized recommendation for users to meet their needs when facing the inquiry service.This paper proposes a method which utilizes review analysis based on natural language processing,extracts feature properties to do the multiple attribute decision making,and finally provides users with recommended ranking.This method established an information recommendation system based on comments semantic and web mining techniques,in order to implement personalized service for customer.The paper has fulfilled the needs of comparison between different shops for the same goods and making reasonable recommendation.We proposed a framework for doing subject extraction and sentiment analysis on user comments on the shops,and then,cluster the result as a Customer Satisfaction attribute.This attribute,together with the objective data collected from the shop pages,was put into the recommendation system for calculation.The system allows user to choose attributes of their interest,and input the importance for each attribute in a sequencing manner.We implemented the recommendation system to give the ranking results which can not only objectively,comprehensively reflect the real situation of the shops,but also can meet the user′s personal preferences.