Font Size: a A A

Design And Implementation Of Multi-Dimensional Personal-based Video Danmu Analysis System

Posted on:2020-04-10Degree:MasterType:Thesis
Country:ChinaCandidate:Y M XuanFull Text:PDF
GTID:2428330572473602Subject:Computer technology
Abstract/Summary:PDF Full Text Request
As an emerging communication carrier,danmu has been accepted by more and more young people.People share their thoughts and opinions through this new communication medium.When users watch a video,they can issue a danmu at any time to create textual interactions with other users.While users issues new danmu,the content corresponding to the sentiment of the current slide of video is often included.More and more researchers are engaged in the research and data mining of danmu.This thesis implements a multi-dimensional persona-based video danmu analysis system.The system is based on Web development and embeds a new character emotion topic model.The topic model is derived from Latent Dirichlet Allocation Topic Model(LDA).It is used to extract the topic words of different characters and different emotional polarities from danmu.The video danmu multi-dimensional analysis system can help researchers to analyze the emotional tendency of the audience more conveniently and quickly,and to extract the key words of the danmu.At the same time,the system also embeds a crawler module to help build a danmu corpus.This thesis first introduces the research background and significance of the topic,and then introduces related technology research.Next,a detailed demand analysis of the persona-based video danmu multi-dimensional analysis system is performed.On the basis of the demand analysis,the persona-based sentimental topic model is designed and implemented.This model can mine fine-grained topic information from characters and sentiment aspect.Next,the multi-dimensional analysis system is designed to outline the system operating environment,system architecture,system interfaces and functional modules.On this basis,the danmu data acquisition module,the danmu data parse module,the persona-based sentimental topic module,the page module and the database construction module are designed and implemented.After that,test each functional module to ensure that the system works properly.Finally,the thesis summarizes the work of this thesis and describes the deficiencies in the system and further work.Summarize my work and achievements during my master's degree.
Keywords/Search Tags:danmu, LDA, topic model, sentimental analysis, persona-based
PDF Full Text Request
Related items