Study On High Utility Review Mining Of Review Community

In recent years,the development of the Internet has reached a new height and big data has been gradually affecting all aspects of the application of the Internet.In the era of Web2.0,users are both content producers and consumers of content,which provides great opportunities for the development of reviews community.However,the explosive growth of data has brought more and more problems: the amount of data in every walk of life has been growing and the on-line comments are uneven in quality.People had a tough time finding genuine and in-depth comments online.Therefore,how to dig out efficiency comments,which can reduce information search costs,and improve the quality of online comments,has become the focus of this paper.The purpose of this paper is to find the efficiency comments,allowing users to spend the least time to obtain the most comprehensive information.There are three main contributions in this paper: Firstly,we established an index system,including classification index,sequencing index and evaluation index.Secondly,we designed a two-layer model to find the useful comments.The first layer is a classification model.Comments will be divided into four categories,which are named high-value comment,potential comment,questionable comment and worthless comment.The second layer is a sequencing model.We sorted comments by using two sorting algorithm.Thirdly,we did some data analysis on dianping.com to verify the effectiveness of the model and finally found that the best sequencing model is based on topic modeling.In addition,this paper also arrives at two conclusions: Firstly,the classification model can help community managers find their current problems and corresponding solutions.Secondly,the sequencing model can give users a better experience.Theoretical innovation of this paper is reflected in four parts,which are the research object,research perspective,research methods and index system.In practice,this two-layer method is more excellent.