Research On The Evaluation Index System Of Chinese Film

With the continuous development of modern information technology,film has gradually developed into an important part of human daily life,which is a new culture and art with a wide audience coverage.In 1994,the reform of the film market began.To this day,the film industry has realized the transformation from the traditional noiseless film to the sound and then to the stereo,convent the single tone to the chromatic colour,and the film box office has also completed the growth from several hundred million to more than 60 billion yuan,which is closely related to the support of the state and the government.With the rapid development of the film industry,movie scoring websites have emerged.With the rapid development of the Internet and media,people have more choices on movie scoring websites.Users can express their own opinions and refer to each other.As time goes by,rating has become an important indicator of film quality,but even so,there are still such situations:some audiences think that the film is good,but the rating is low;some audiences think that the film is not so good,but the rating is very high,which shows that it is too one-sided to judge the quality of the film simply from the rating.Therefore,this paper considers many indicators,from the audience rating and the basic information of the film,to establish the Chinese film evaluation index system,so that the audience can easily find their own films.At the same time,understanding many factors that affect the film evaluation in China can improve the economic benefits of the film industry,which plays an important role in economic and cultural progress.This paper makes a systematic analysis of the domestic and foreign movie rating websites,and then determines the data to be selected for the next research.The data are from the top 300 films in Chinese mainland recently,and the cut-off date is December 30,2019.Through data preprocessing,the final sample set is 272 films,and the data is divided into two data sets:the basic information of the films and the score records of film reviews.This paper first determines three objective indicators in the evaluation system,namely,director effect,star effect and film type.Secondly,we analyze the difference of movie score,and find that when the difference of movie score is very large,that is,when the variance is very large,the value of movie score,that is,the average,is very limited.When we use k-means algorithm to classify the samples,we find that there is a mixed distribution in addition to the five common scoring types.Therefore,according to the characteristics of the data,and through the EM test,the final selection of the two components of the mixed binomial distribution fitting,the selection of fitting indicators are the mixed proportion,the standard advantage value of the positive sub population,and the standard advantage value of the negative sub population.Finally,this paper uses entropy method to give weight to the six selected indexes,explore the importance of each index,and establish the Chinese film evaluation index system.The empirical research results show that movie score can reflect the public's love for movies to a certain extent,but it only represents the average view of the audience.When there are great differences in people's views,the value of the average score is very limited,and the movie score can not cover the internal characteristics of the data,so as to measure the film quality slightly single.In the first level index of film evaluation index system,the weight of fitting index is greater than that of objective index.From high to low,the six index weights are respectively mixed proportion,director effect,star effect,standard advantage value of poor evaluation sub population and standard advantage value of good evaluation sub population.When we divide comments into positive comments and negative comments,the influence of negative comments is greater than that of positive comments.And the director and actor of the film have a certain influence in the evaluation system,but the weight of the film type is very low,and the film type has little influence on the evaluation of the film.The main innovations of this paper are as follows:first,the hybrid model is combined with the reality,and the two-component hybrid binomial distribution is used to fit the movie scoring data,which makes up the blank of the hybrid model in the movie scoring research.The second is the combination of objective index and fitting index.The diversity of index makes the final index system more concrete and comprehensive.