Research On The Relationship Between Films'Online Word Of Mouth,Rating And Box Office Based On Text Mining

With the continuous development of the Internet,the research of online word-of-mouth has been deepened.In practice,more and more consumers look up the online word-of-mouth to help them complete the consumption decision.In the film industry,the audience has the habit of writing film reviews and sharing it on the Internet.The length and quality of the text are satisfying.The review,as the reputation of the consumer to the film,is the ideal object of the online word-of-mouth analysis.At the same time,for film works,high box office performance is frequent,but the movie still calls for a good reception and applause.To solve this problem,we extract the characteristics of movie online word of mouth through text mining.We use the results of word of mouth text mining to complete the classification of movie ratings and box office,and build the relationship between film reputation and movie score and box office.In this paper,first of all,the crawler gets more than 140 thousand movie reviews,and combines the word2vec algorithm with the professional knowledge in the field,and proposes a way of building a word table with a better efficiency than a complete reliance on expert knowledge.Secondly,using the constructed keyword table to score the word segmentation after the participle processing,complete the quantitative conversion of three aspects of the word of mouth text in the quantity,validity and emotional arousal of the audience,and make the variables constructed as the characteristics of the machine learning,and verify the score of the film and the performance of the ticket room.Finally,using the characteristics of word of mouth extract from natural language processing,the quantitative relationship among word-of-mouth text,movie score and box office is constructed.The results show that,first,the variables of word of mouth constructed by text mining can effectively predict the score of the film and box office performance compared with the frame in the past.Secondly,the volume of shot and the emotional arouse positively influence film rate.The volume of actor and the negative valance of story negatively influence film rating.The volume of story and emotional arouse and the negative valance of actor positively influence film box office.The research of online word-of-mouth,combed with machine learning algorithm and word vector coding,which reduces the difficulty of text mining in the professional field.It provides a positive exploration for the research of word of mouth content analysis,and broadens the dimension of word-of-mouth research.The relationship among the movie box office,the rates and online word of mouth is constructed.