Research And Implementation Of Mobile Game Comment Public Opinion Monitoring System Based On Text Tendency Analysis

With the rapid growth of the mobile Internet and the large-scale popularization of smart terminal devices,mobile games have become a mainstream way for us to entertain in fragmented time.The major mobile application markets provide players with convenient mobile game download channels.At the same time,players can also easily comment,complain or provide suggestions on the mobile games they play on game communities,Weibo,forums,and application markets.How to further dig out the commercial value of a game requires game industry personnel to always put the player's playing experience in the most important position.Therefore,in the development,testing and operation stages of the game,the game industry personnel are eager to know the reputation of the game among players,and use this as a guide to make high-quality games that meet the tastes of the public.For game operators,these comments reflect the new version of the game and the recent operating activities' response and word of mouth among players.By observing these valuable comments,they can obtain the optimization policy of the game and operating activities.On the other hand,for the application market,ratings and reviews data can also be used to dynamically adjust the game rankings of the leaderboards.Based on the field of mobile games,this article researches the pivotal techniques of sentiment analysis of massive mobile game reviews.Based on these key technologies,this paper uses the Python language,through My SQL,Django and other tools and frameworks,to implement a set of comments and public opinion monitoring systems in the field of mobile games.The main work of this paper is as follows:(1)New word discovery and emotional dictionary construction in the field of mobile games.By crawling a large amount of mobile game review data,this paper obtains a sufficient amount of review corpus,and builds a dictionary in the field of mobile games through the method of new word discovery.After acquiring the dictionary,the recognition of emotion words and the classification of positive and negative emotion words are further carried out.(2)Emotional orientation analysis of mobile game reviews.By constructing LSTM neural network and Bi LSTM neural network models to analyze the sentiment orientation of reviews,the positive and negative emotions hidden in the mobile game review can be effectively distinguished by this model.At the same time,the mobile game sentiment dictionary constructed in this paper is used for feature extraction,and the experimental results obtained are compared with the current more mature sentiment classification methods.(3)Design and implementation of mobile game comment public opinion monitoring system.Based on the above research foundation,Django,which is widely praised by developers,is used as the web framework to build a set of review and public opinion monitoring system in the field of mobile games.