Opinion Mining And Ranking Based On APP Review

With the popularity of the Internet,people increasingly like to express their views on the Internet,which in turn has become an important reference in the people's daily life.The field of mobile APP is no exception.The APP review on the mobile market is a major factor influencing people to evaluate and download APP,and affect the mobile Internet traffic level.How to mine important information from a large number of APP reviews and prioritize key reviews has become a pressing issue.Currently there have been some research for reviews on the e-commerce product,but in the field of app reviews we do not found any.In this paper,we take the APP reviews from a large-scale APP download plat-form in our country as the research object,and research a comment mining algorithm that is suitable for the opinion mining and sorting on APP reviews.The algorithm is based on the existing research algorithms of e-commerce shopping and living informa-tion reviews in academia.The algorithm focuses on user-friendly and comprehensive information about APP reviews.The main contribution of this article has three points:1.It analyzes the shortcomings and miscellaneous APP comments,proposes the data cleaning specification required for APP reviews,and puts forward the re-quired indicators for APP reviews spam.2.This paper proposes a opinion mining algorithm based on semantic clustering,and based on this,we excavate the opinion of APP review data.Experiments show that compared with the traditional opinion mining algorithm,this algorithm effectively improves the result.3.On the basis of opinion mining,this paper studies the index of the authenticity of the comment,and then puts forward a APP ranking model that combines the quality and authenticity of the comment.Compared with the traditional ranking model,the APP can better grasp the viewpoints of the whole comment.