Research On Online Impact Factors Of Customer Reviews Usefulness Based On Book Reviews Data

As the development of e-commerce,online shopping becomes more and more poplular beacause of its advantages of convenient,fast and so more.As the evaluation of other cusumers,online reviews are very important information sorce in consumers making purchasing decisions.However,facing lots of online reviews on the internet,how to select the effective information fast becomes a concern. This paper takes books'online customer reviews as the subject,uses lingpipe analysis and statistical method,studies the reviews'helpfulness impact factors from the view of text features.This paper describes the online reviews and consumer psychology-related concepts and theories first,analysis the impact of online review on consumer decision making from the view of information dissemination process and consumer process information.And then analysis the impact of different polarity of reviews on consumer perception,according to attribution theory and emotional coherence theory.Then this paper analysis online reviews'helpfulness impact factors,presents the assumption of impact of each factor on reviews'helpfulness,and establish a modle of online reivews'helpfulness impact factors.We crawl data,extract the value of text feature,analysis model result,we conclude that online reviews with the feature of positive,subjective and late published are more useful.In order to indentify the usefulness of a reivew,we predict the value of review usefulness by the model,and assess the effect of predition.In the end,we analysis the reason why reviews lack of helpfulness votes,establish the logistic model from the view of reviews'feature.After crawling review data and analysising the result of regression,we conclude that the reviews that is too subjective,title of which is too positive or the publish time is too late will be lack of helpfulness easier.Reviewers should pay much attention to these factor to make their reivew more effective.We can predict the helpfulness of a book review useding the model established in this paper to sift the reviews of high helpfulness,it can help the consumer to improve their decision-making efficiency.Moreover,the paper analysis the reason why review lack of helpfulness vote.We believe that this paper will be useful in the future.