Font Size: a A A

Design And Implementation Of Visualization System For Movie Website Data Mining

Posted on:2020-10-24Degree:MasterType:Thesis
Country:ChinaCandidate:Z WangFull Text:PDF
GTID:2428330590950615Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of the Internet industry and the film industry,the links between the two are becoming more and more close.Many Internet video websites have emerged.Traditional Internet video websites have gathered a large number of movie resources,providing users only to watch and download.Channels,and users want to find the right movie to watch in a large number of movie video resources is very difficult,in order to give users a certain reference in making objective and reasonable choices in the massive movie video resources,from the movie network film as the entry point design Visualization system for film data mining.Because the film reviews on the web contain the rich feelings and tendencies of the viewers,they also include the degree of association between different movies on the emotional and semantic levels.The system uses the film review text as the entry point of the study,and explores the connection between movies in the text.Firstly,a set of movierelated data based on Scrapy lightweight crawler framework was collected to provide extensive and reliable data support for the entire visualization system.Based on the completion of the film review text,the word segmentation was used to stop.Using words,constructing a text preprocessing method such as an emotional dictionary in the exclusive field of the film,and simultaneously using the constructed emotional dictionary to analyze the emotions of the film review text,and secondly,extracting the features of the document composed of the film review text and the movie profile,and then using the basis The distance clustering algorithm K-Means classifies the captured movies.Since the keywords of the film review are required to be displayed in the form of a word cloud,the commonly used algorithm TF-IDF for keyword extraction is used to extract the keywords of the film review.Finally,in addition to an analysis of the film review text,the Echarts plug-in was used to make a visual display of the statistical analysis of the film review user and the film review operation.Based on the use of the system,it can help users to obtain the overall emotional tendency of the movie review text,and organically combine the personalized service with the wisdom of the public,which can truly reflect the feelings of different users on the movie,which can satisfy User personalized requirements,can effectively monitor social sentiment and provide a more reasonable and objective reference for users to watch.
Keywords/Search Tags:data mining, data crawling, preprocessing, data visualization
PDF Full Text Request
Related items