Analysis And Prediction Of Factors Influencing Chinese Film Box Office

In recent years,with the rapid development of China's film industry,the revenue of movie box office in China has exceeded 55.9 billion yuan in 2017.However,film investment is still a high-risk and high-tech investment game.Therefore,the study of the factors that influence the box office of the movie will help the movie party to provide the basis for film investment decisions.At the same time,with the advent of the Internet and the era of big data,the factors that influence the country's movie box office have also undergone major changes.By combing domestic and foreign movie box office related research literature,this article combines the development status of China's film market,finds out the influence factors of Chinese movie box office and establishes an index system,and then predicts Chinese movie box office by constructing a random forest model,so as to be able to provide movie box office research New research methods and ideas.This article selects domestically produced movies(excluding cartoons and documentaries)with a total sales volume of more than 50 million at the box office in 2015-2017.A total of 170 titles are used as research objects.This paper reviews domestic and international literature on the factors that influence the box office,and based on the development of the Chinese film market,from film production.In terms of marketing,distribution and distribution,25 variables were selected as the influencing factors for the box office of Chinese movies.The research data comes from Douban.com,China Film.com,China Entertainment Index.com and other official websites.It uses Python crawler technology,calls related APIs,and writes python programs to collect data.This article first describes the statistical analysis of the distribution of domestic variables at the box office.Second,using multiple linear regression and random forest models,explore the influencing factors of high-priced box-office movies.The empirical process uses SPSS software,and finally establishes a weighted least squares regression model,and draws the rowing rate,number of ratings,release period(New Year file),celebrity appeal,point projection box office,movie type(adventure/science fiction),movie type(action(/Crime),release period(other period),story familiarity(sequel),movie type(war),these 10 variables have a significant impact on movie box office.Using python to establish a random forest regression to get the importance of each variable,the results obtained that the top ten variables of importance are:point to the box office,want to see the number of people,ratings,platooning rate,movie score,celebrity appeal,favorable rate,bad ratings,celebrity influence.Finally,using the model to predict the box office of 9 domestic films(except cartoons,documentaries)with a box office of over 50 million at 2018(as of March 2),and comparing the prediction effect of the two models,the results show that the random forest Has higher fitting accuracy.The results of the study showed that the filming rate,number of ratings,screening period(Lunar New Year file),celebrity appeal,point-to-box box office,movie type(adventure/science fiction),movie type(action/crime),release period(other period),and familiarity with the story The degree(sequel)and the type of movie(war)have a significant positive effect on movie box office.Based on the research conclusions,this article proposes related recommendations for the development of domestic films,and has a strong practical significance for domestic film investors and decision-makers in theaters.