Research On Recognition Of Film False Comment

In recent years,Chinese movie industry has continued to develop healthily,and the number of online movie reviews has increased rapidly.Many consumers use movie reviews as a reference when choosing movies.Facing the temptation of huge interests,many film teams released false film reviews through the purchase of the Marines.While continuously improving the level of film reviews of their own films,they also attacked other films in the same period.False movie reviews produced by such unfair competition methods need to be taken effective measures to be managed in time.The main reasons are as follows:First,consumers will be misled by false movie reviews,weakening their trust in online movie reviews;Second,a large number of film teams use resources for one-sided film criticism promotion,while ignoring film production,resulting in the continuous decline in the quality of film works.In view of this,it is particularly critical to effectively identify and timely manage the false reviews of movie platforms.As far as online movie reviews are concerned,the review information is the most representative and an effective data source that can be used to identify false reviews.This article uses it as a data set for research and analysis.This article comprehensively analyzes the research status of false comment recognition and semi-supervised learning,details the principle and specific classification of semi-supervised learning,and finally uses the Tri-Training algorithm of semi-supervised learning for false comment recognition,and compares it with the commonly used in the past.The supervised learning support vector machine classifier is compared and analyzed,and it is concluded that introducing unlabeled data into the classifier training process is theoretically feasible,and it can also perform prediction and classification tasks in practice.It also shows that it is feasible to introduce semi-supervised learning into the field of false comment recognition.