| The epidemic is sweeping across the globe,causing a major impact on the global economy and society,and posing an unprecedented challenge to the "One Belt,One Road" initiatives.With the development of online technology,social media has increasingly become an important platform for citizens to share their personal feelings during the epidemic prevention and control period.Controlling public opinion during the epidemic,promoting public information and scientific action to combat the epidemic,and reducing the impact of undesirable and false information is an urgent need for the country in emergency situations,and a forward-looking way to deal with the major changes in international relations in the aftermath of the epidemic.In this paper,the English-language texts of the Six Corridors and Six Roads of the overseas social media platform Twitter were selected as the target of the study,and Asian countries along the "One Belt,One Road" cooperation agreement with China were taken as the main subject of the study to explore public opinion and analyse the spatial and temporal patterns of the new crown epidemic in the region.The main research work and results are as follows:(1)A framework of social media text theme extraction and sentiment classificationThe extraction and sentiment classification framework of social media texts related to the New Coronary Pneumonia incident in the six corridors and six roads.Based on natural language processing technology,the text information was preprocessed with punctuation,numbers and URLs.A topic extraction and classification framework was constructed based on topic models and machine learning methods to extract topics from Twitter users in the context of the Newcastle pneumonia incident.The BTM topic model combined with the random forest algorithm was used to achieve topic classification of 15 sub-categories and 6 major categories.The text data was subjected to topic class identification to each topic assigned to its closest topic,and the text topics and topic distribution of Twitter users in the social media of the New Crown pneumonia event were obtained.(2)Research on social media text geolocation information framework in the six corridors and six roads regionBased on the named entity and global geographical location parsing framework,the geographic location of public topics in the six corridors and six roads was obtained from the social media texts related to Newcastle pneumonia and their spatial distribution pattern characteristics were mined.The Spa Cy NER tool was used to extract countries or regions from Twitter texts,and Geo Names was used as the data source to match geographical names from multiple sources of global geographical names data,effectively solving the consistency judgment of geographical name information of text data and realizing the organic fusion of multiple sources of global geographical names database.It also discovers the evolutionary characteristics of themes and topics in different regions under five scales of Central Asia,East Asia,South Asia,Asia,West Asia and CIS,and explains the spatial differences between themes and topics.(3)Research on spatio-temporal mining analysis of social media topics in six corridors and six roads regionCombining the accumulated confirmed cases and deaths,population density,GDP per capita,and export value of goods and services,we analyze the relationship between the evolution of public opinion in 46 countries in the Six Corridors and Six Roads region under five regions of Central Asia,East Asia,South Asia,Asia,West Asia,and CIS,discover the evolution pattern and spatial differences of regional themes,explain the reasons of theme distribution,and provide public opinion analysis for epidemic prevention and control and international cooperation in the Six Corridors and Six Roads region.It provides data and methodological support for epidemic prevention and control and international cooperation in the six corridors and six routes. |
PDF Full Text Request