Analysis On The Semantic Correlation Of Social Network Public Opinion For One Belt And One Road

In the era of big data,how to quickly and efficiently analyze massive amounts of multi-source data and establish public opinion monitoring and guidance mechanisms to provide decision support for managers is a hot and difficult topic in current research.With the rapid development of modern technology such as spatial data acquisition,storage and analysis and processing.On the one hand,the volume of spatial data expands dramatically,and the complexity increases significantly.On the other hand,many important spatial information hidden in spatial databases and their associated non-spatial data are not fully utilized.Especially for network public opinion data analysis,development is slow.Sina micro-blog has tens of millions of users in China,and produces a large number of user information every day.It is of high application value to analyze and excavate Sina's micro-blog public opinion data.This paper focuses on the analysis of spatial semantic relevance analysis of social networks in the “One Belt and One Road” under the premise of the current research on the Sina Weibo network.Using sina weibo network public opinion data,through natural language processing and spatial analysis methods.The main contents are as follows:(1)Through the analysis of Sina Weibo's public opinion data on the “One Belt and One Road” topic,the key words for data acquisition are set,and two methods of data crawler and the data interface provided by Sina Weibo Open Platform are used to obtain Sina Weibo Networks.Public sentiment data.(2)This paper analyzes the acquired Sina micro-blog network public opinion data,designs the data storage type and data organization form of the network public opinion,and stores the data using the MySQL network database,constructs the main key and the database engine,so that the data retrieval and other operations are more convenient.(3)A suitable method for obtaining geospatial data is selected,including the address data and IP address resolution of the post text,and the latitude and longitude coordinates of the Baidu map coordinate system are obtained through the Baidu map address resolution API,and the coordinate of the block is acquired through coordinate system conversion.Through Chinese text segmentation technology for word segmentation processing of post text,word frequency statistics blog hot word,it is concluded that the high frequency words,and through the analysis of sina weibo topic forms,combined with high frequency words summed up the hot topic.(4)The text semantic similarity classification model is constructed,and the corresponding topic category is automatically divided into the unknown blog text.By constructing semantic tree,the semantic relational degree of the topic is calculated,and the correlation degree between each node of semantic tree is obtained,which provides support for information retrieval.The temporal and spatial characteristics of online public opinion are analyzed,and the changes of online public opinion according to time and space are summarized.For Baidu map tile service,spatial grid semantic clustering analysis based on tile grid is implemented.(5)Combining related technologies of WebGIS,using Vue,Laravel and other development frameworks,build a B/S-oriented “One Belt and One Road” social network public opinion spatial semantic correlation analysis and visualization platform.