Research On The Model And Application Of Public Emotional Network Communication Based On Machine Learning

With the Internet development,social platforms have emerged,with the explosive growth of data on the network,among which text data is a representative form of data.Text data often carries personal emotional color,so emotional analysis of text data is an effective means to understand social public opinion tendency,product word-of-mouth,quality of service and other issues.Not only that,it closely links the masses as an invisible network,and the various emotions of the masses are constantly exchanged with this network as a carrier.Faced with such huge text data and complex mass relations,it also poses new challenges for industries in various fields.Emotion analysis has always been a hot trend in various fields.Based on this,this paper puts forward a new research method for emotions in the field of logistics on the basis of predecessors,proposes the naive Bayesian classification algorithm of machine learning for emotional text set of logistics,analyses the emotional tendency of the masses,and on this basis,constructs various emotional networks of the masses,and establishes the emotional network of the masses.Characteristics and propagation paths are studied.Firstly,this paper cleans and preprocesses the crawled emotional texts related to logistics.After removing the useless duplicate texts,the text data is divided into training data and test data.In order to make the computer recognize the text content,the emotional text is segmenting and removed.The first two hundred words are selected as feature words by word frequency statistics.After calculating the TF-IDF value of feature words,the Naive Bayesian classification model is trained.Then the text data are classified into positive,neutral and negative emotions by the trained model.It is found that the majority of people's emotions are positive,the second is neutral and the least is negative.On this basis,using Ucinet6 software,using users as nodes and forwarding relationship as the boundary,the overall emotional network and various types of emotional network of the people who publish emotional texts are constructed.After studying the communication characteristics and paths of various types of emotional networks,it is found that the relationship between users of negative emotional networks is the closest.Negative emotions are more likely to arouse public resonance.The forwarding amount of users of positive emotional networks is mostly between 0 and 1.There is no positive emotional leader or neutral emotional network.There are many focuses and need to be guided by positive emotions.Emotional information in the network began to spread in a"central" way to the four-level,in the process of dissemination may produce secondary user nodes,and then to the two-level user-centered nodes to spread around,and then form three-level,four-level user nodes,until the information is transmitted to the public with zero node degree.Finally,according to the case of emotional analysis and emotional network research results,summarized the current logistics industry problems.Let managers realize the role of big data in promoting the industry,and propose to make use of the functional characteristics of big data,improve the means of logistics service monitoring,and optimize the management environment of logistics terminal network.In the optimized emotional network,the company disseminates the improved service content,improves the user stickiness and increases the number of users.With the support of big data,we should optimize the delivery route and improve the efficiency of distribution.