Research On Event Evolution Mechanism And Trend Prediction Method Of Social Network

Social computing is a interdisciplinary field.Social computing has entered a phase of rapid development with academics and industry pay more attention on this field and the level of recognition is gradually growing.The related research and applications continue to emerge.Different from the traditional information media,the form of the information in the social network platform is more complex,and the communication process is stochastic and paroxysmal.In this paper,we try to extract various features from the actual public data of the Internet to predicting the event trends in a quantifiable way.And finally,predict the trend of events from a macro perspective.Based on the analysis of the characteristics of information dissemination on social network platforms,we propose a genetic information network,and according to the characteristics of social network platform,provides a method to predict the genetic events based on the trend of network information,network information based on genetic growth rules and events trend is predicted for an improved population model put forward.Finally,expand the event trend forecast factors,proposed a theory of Clausal Pivot sentiment analysis method based on social network information platform for sentiment analysis,and the method of quantitative analysis for event prediction.The research work is summarized as follows:(1)Research on the information dissemination mechanism of social network platform event evolution.In a social network event evolution pattern is a pattern,but also contains the user data in the event of the evolution of relational data,so this paper directly from the event evolution model,carries on the analysis to the social network and obtain useful data,the biological evolution theory,introducing the ideas for the analysis of event evolution characteristics,firstly,the information dissemination process the example analysis,then design a genetic information network,at the same time to determine the event evolution theory paradigm,and discusses its use scope.It provides a new way for the analysis of information dissemination in social networks.Based on the information network,this paper presents a framework for event trend prediction based on information network.This paper uses the related time series prediction to predict the trend of events,this paper analyses the factors influencing the evolution trend of events,the prediction method for time series prediction problem solving event trend to verify the effectiveness of the method.(2)According to the genetic characteristics of information,it is the goal of this paper to improve the accuracy of trend prediction in the framework of event trend prediction based on information network.Considering the biological characteristics of the genetic information of the network,the paper proposed a prediction method of population improvement trend based on the event model,the model is mapped to an input for the n parameter,BP neural network output for the 1 parameters,which influence the trend of events,n said a number of factors,combined with the impact of the trend of events to predict the trend of events.(3)On the dimension of influencing factors,we consider the emotional polarity of information,and adopt a sentiment classification method based on sentence structure to determine the emotion of the individual.In cascaded conditional random fields model,through training clause fine-grained conditional random field emotion classification model,the random condition of coarse grain field training sentence sentiment classification model,finally through the cascaded conditional random field model for information content of individual sentiment in classification,and the polarity of quantized input as the influence of a prediction model of the dimensions of input factors,enhance the accuracy of incident trend prediction.