Research On Identification Of False Movie Critics Based On Logistic Regression

In recent years,with the continuous development of the Chinese film market,online film reviews have become an important evaluation indicator for consumers' viewing choices increasingly.However,under the huge commercial interest,many false reviewers are not only using false film critics to increase the evaluation of the target film deliberately,but also discredit the competition films in the same period arbitrarily.If these false reviewers disrupting the normal user's viewing choices cannot be stopped effectively,they will continue to mislead consumers who watch movies and inflict their trust in online film critics;on the other hand,more and more film and television teams will ignore their film and television production,and put too much money and effort into such bad mistake.At present,there are certain developments in the research of false commentary recognition in the field of e-commerce.Its main research direction is to start with two aspects.The first way is based on the content analysis,and the second is based on the false reviewers' attributes and analysis of behavioral characteristics.However,the research on recognition of false film criticism in the online film review environment is still very limited.At the same time,the false reviewers in the website of Douban is different from the false reviewers in other fields.Its identification research has particularities:the data of the false reviewers content and behavior show the“normal”.This "normalization" is due to the fact that the false reviewers groups have adopted"Account maintenance" and recruited normal users.Therefore,the false identification methods based on false comment content or reviewers' behavior characteristics in the past have limited effectiveness in the identification of fake film critics.In the research of false criticism of this article,based on the relevant research and the characteristics of the false reviewers group in Douban,a solution was proposed and a false film critics'prediction model was constructed.The research goal of this paper is to explore how to identify the false reviewers in the online film review.The specific content is to construct a false film critics recognition model based on logistic regression with date of Douban.The research firstly combed the relevant researches of false commentary identification and online film review in the other e-commerce fields.Based on previous researches,we set a specific target set index for realizing the characteristics of false reviewers.Based on the data of the characteristic indicators,a predictive model of false critics was constructed by using logistic regression,and the effectiveness of the extracted indicators was verified.Also this study lays a solid foundation for the follow-up research on recognition of false film reviews.