Research On The Construction Of A Model Of World Political Conflict Based On Big Data

Conflict prediction has always been a subject of great concern in the international political field,but traditional international conflict event data is scarce,and fine-grained observation data is difficult to collect on a large scale.Therefore,predictions have obvious shortcomings in terms of accuracy and timeliness.The emergence of an event database based on massive big data has greatly improved the data volume and quality of international conflict research.At the same time,increasingly powerful data analysis technologies have made it possible to observe and analyze political violence on a large scale and explore conflict dynamics,as well as the possibility of conflict prediction.The overall goal of this paper is to use a large and fine-grained Global Database of Events,Language,and Tone Database(GDELT),conduct detailed empirical analysis through actual conflict cases,and build a corresponding model using LSTM RNN based on the event big data provided by GDELT.Conflict dynamics for predictive warning.At the same time,given that publicly accessible Internet social networking sites(e.g.social media)play a key role in information dissemination,this article selects two cases,through corresponding data capture and analysis,to verify the public's reflection through social media The impact of attitudes on the trend of political events,that is,the effective monitoring of the public's attitudes and emotions towards political events can help reveal public opinion trends that affect social and political instability.Therefore,this article proposes that the prediction of international conflicts should be based on the GDELT event big data model to make the first judgment,and social media big data should be used to analyze the public's emotional polarity to the event.,As the second criterion for judging whether a conflict occurs,so as to improve the accuracy of conflict prediction and early warning.The paper consists of seven parts.The first part is the introduction.Firstly,the background of the topic selection,the topic selection,the research question,the research significance and the data source are introduced.Then the concepts related to this research are defined and analyzed,including big data,big data analysis,and computing Relations,etc.;then review the current status,path,and problems of domestic and foreign research;review the innovations,difficulties,research methods,and basic structure of the research.The first chapter focuses on the theoretical elaboration,which involves the theory of international conflict and the data basis of this study.It defines the concepts related to international conflict,discusses the development of international conflict and peace research theory,and the current situation and shortcomings of conflict prediction research.This leads to the data foundation of this study-big event data based on the GDELT event database,and the reason for choosing this data source for conflict prediction.The second chapter selects the “Arab Spring”,the conflict in Sudan from 2018 to 2019,and the “yellow vest” incident in France as examples.Based on the classification of the event data(verbal cooperation,material cooperation,verbal conflict,and material conflict),the above cases were analyzed based on the data provided by the GDELT database.The results show that a significant increase in the proportion of verbal and material conflicts in the total number of incidents can be a sign of an outbreak of conflict or increased tension.The third chapter starts with big data of social media.Under the situation that social media is increasingly affecting citizen participation and political processes,it will monitor the public's attitudes and emotional changes to political events through the mining and analysis of social media data.This chapter selects two cases of Brexit and impeachment of Trump,confirming that social media data can reflect the influence of people on political events,and then help predict the trend of political events,including political conflicts.It fully demonstrates that social media data can be a useful supplement to the research on big event data prediction derived from mainstream news media.The fourth chapter constructs the conflict prediction model.It details the basic assumptions of the model,the algorithms and techniques used,and the specific construction process.Then test and verify the conflict prediction model that has been constructed.The prediction results show that the conflict model based on the LSTM RNN algorithm can much more effectively predict the occurrence of related conflicts and provide early warning of possible situation deterioration.The fifth chapter discusses the risks of big data application in the field of conflict prediction,and the problems and challenges of conflict prediction based on big data.The conclusion section summarizes this research and puts forward the future research directions.