Judgment And Classification Of Media Negative Vocabulary Based On AlphaGo Double Brain Mechanism

With the rapid development of online media,internet users use the network to communicate which makes the network language develop rapidly.It has gradually become an indispensable part of the people,and has become the research object of researchers.It is undeniable that the development of online language has greatly enriched the expression of Chinese language and promoted the development of Chinese language and characters.At the same time,it should be noted that the network language is full of irregular,discordant and unhealthy content,such as,typos in the language of the Internet,language violence,insults,pornography and other issues.Therefore,how to correctly analyze the problems existing in the network language and detect and track the network language problem in time has become an important practical problem,and brings new challenges and new topics to the researchers of the network media language.With the continuous development of artificial intelligence,breakthroughs have been made in many fields today.In 2016,AlphaGo officially appeared in the public.It defeated world champions Li Shishi and Ke Jie in the Go game,which made the world sensational and gained wide attention.The emergence of AlphaGo has also made the technology of artificial intelligence gain the attention of the world,making artificial intelligence research the focus.It consists mainly of two neural networks,one is a strategic network and the other is a value network.These two "brains" work together to improve chess.Inspired by AlphaGo's trategic thinking,this paper identifies and classifies bad vocabulary for purifying the network environment.Based on this research topic,the main work of this paper has the following points:(1)Study the text segmentation method.The primary task of the identification of bad vocabulary is to divide the sentence into words.The word is the object of this study.After obtaining one word,further research work can be carried out.(2)Construction of bad word lexicon,constructing bad lexicon,and processing bad vocabulary into the library.Through the structure of the thesaurus,the detection and segmentation of bad words can be made more accurate,which lays a foundation for subsequent text processing.(3)The determination and classification of bad words,the identification and classification of bad words is the key work of this paper.The research of these two works is used to achieve the purpose of this paper.(4)The construction of dual decision models.The identification and classification of bad information is operated by two "brains".One brain is the reptile to obtain text and extract features and then information identification,which is the same as AlphaGo's"drop selection" network;The brain is based on rules and statistics to compare and classify bad information bases,just like AIphaGo's "value network."...