Research On The Algorithm Of Spam Information Processing In Social Network

Nowadays,with the development of Web2.0,Social networks play an increasingly important role in people's lives.Now there are many the popular social network platforms,such as Sina Microblog,Baidu knows,WeChat,QQ,Live app,Zhihu,Douban,etc.At the same time,with the popularity of mobile phones and other communication tools,it provides a convenient for people to participate the online activities like surfing,sharing information,communicating with each other at anytime and anywhere.However,a large number of spammers have been generated.These users push malicious links,promote fake advertising,asperse others and spread rumors,etc.on the platform.This behavior has seriously affected the user experience,brought trouble to people's lives,getting more and more serious negative impact.Therefore,how to identify and detect these malicious users and shield spam become the current research hot issues.This paper chooses the mainstream social platform like Sina microblog and the knowledge sharing platform Baidu knows,using machine learning technology and sorting ideas respectively on the two platforms to deal with spam.We design the spammer detection algorithm for microblog and an acceleration falling algorithm for answer ranking for Baidu Knows.The main contents of this paper are as follows:First of all,we introduce the definition and the development of social networks and the common network of spam problems for microblog and Q & A website respectively.Also a summary of the garbage problem,including spam classification,processing technology.Secondly,in this paper,we propose a spam detection algorithm based on color classification and a spam content detection algorithm based on blacklist of words.At the same time,based on these two feature sets,we propose a method based on Bayesian network Spam detection algorithm.Experiments show that the Bayesian network based spam algorithm is better than the naive Bayesian algorithm,and it is better than the algorithm for spam behavior and spam content detection.Finally,for the spam information in the Q & A website,the spam answers are divided into the explicit one and the implicit one.And the answer to the implicit spam answer,which is more difficult to use technical means to detect,is presented.The idea of object falling process is cited.Prove that the model can effectively trash the answer to the bottom of the answer sequence.