Research And Implementation Of Emotion Analysis System Of Network News Comments

With the progress of Internet technology,the way of people's living,thinking and behavior has been influenced tremendously.People increasingly inclined to use the mobile internet to get some information.It makes numerous media shifting from offline to online.More and more news corporation,such as NetEase News,Tencent news and Baidu News,successively development mobile news client.Netnews has gradually become the main way to getting news.With the enhancement of the Network interactive ability,the netizens can express their own opinion and emotion through the comments.In this process,the network news opinion will be apper.In order to understand the emotional tendencies of netizens in time,the Netnews-oriented comments analysis has great significance.Therefore,this paper investigates an Netnews comments analysis system to timely monitor the general tendency of the netizens' attitude about some news event such as policy,social hot events and so on.This system plays an auxiliary role in emergency management of social institutions and the government,it help them promptly grasp the trend of public opinion,perceiving the impact of the public event as soon as possible and take some corresponding measures and respond appropriately.Through the study of the characteristics of news dissemination,it has been found that the online news is updated at a faster rate.The number of netizens participating in the review is higher.Using these comments data to analysis emotional can help us judging the emotional tendencies of real time news and hot events on the Internet,monitoring the trend of public opinion more effectively and providing a reference for the guidance of the network public opinion.The main functional modules of this system include news comments obtaining,news data retrieval,emotional analysis and data visualization.In the news comments obtaining module,it contains the management of data template,news data obtaining and comments data obtaining.The news data retrieval module mainly contains time retrieval,keyword search and news source retrieval.The emotional analysis module contains the construction of theemotional dictionary and the comments relational tree,and emotional analysis.In the data visualization module,we can visualize data using line chart,histogram and pie chart.In this system,it present an synthesis algorithm which based on two existing emotional dictionary HowNet and NTUSD to construct the emotional dictionary.The system also put forward a comment relationship determination rules in the process of the construction of the comments relational tree.Meanwhile,analyzing the ultimate emotional tendency of comments using commentary emotion derivation rules.The purpose of these operations is to improve the accuracy and efficiency of comments' emotional tendency.Through the system design and implementation,the system is basically able to accurately analyze the emotional tendencies of comments,achieve visualization of the emotional tendency of news,and intuitively identify the distribution and trend of the emotional tendency of netizens.