Research On The Hot Topics And Urban Differences Of Smart City Based On Text Mining

The development of smart city technology and projects is one of China's key national policies,and also an effective method of solving urban problems.With the gradual landing of smart city projects and the increasingly diversified of smart applications,it has become even more important to strengthen the monitoring and understanding of smart city development and to provide accurate data-based guidance.So far,research on China's smart city has focused mostly on the qualitative theoretical level,whereas research on the quantitative empirical level is relatively few,facing the bottleneck of insufficient data support.As a result,research on hot topics and urban differences of China' s smart city is quite limited.In this context,this paper evaluates the development of China' s smart city by conducting empirical research on its hot topics and urban differences,so as to better understand its developing status,identify existing problems and make suggestions on the development of smart city.Meanwhile,this paper discusses and summarizes the advantages and applications of its data and research methods based on their application in this research,so as to promote the use of new data and text mining technology in urban research.In empirical research,the analysis results can be concluded from three aspects.(1)In terms of hot topics,the degree of attention to China' s smart city has kept increasing in recent years.The contents of attention are mostly on macro scale.Different applications have different hot topics:smart tourism focuses on regulation,smart transportation focuses on science and technology,smart logistics focuses on e-commerce,smart medical and smart community focuses on services,and smart education focuses on informatization.(2)In terms of contents,the development of China's smart city prioritizes traveler experience over local resident service and city management.Smart applications have relevance with one another and can be classified into three categories and eight subclasses.The contents of different applications reveal the core of their development,for example,tourism should provide a more comprehensive tour guide service,transportation should improve infrastructure construction and promote unmanned driving,logistics should pay attention to customer experience and delivery,medical care should improve the efficiency of one-stop medical service,community should concentrate on the essence and purpose of service,and education should focus on smart teaching methods.(3)In terms of urban differences,the correlation degree between urban and smart city shows a long-tailed characteristic,which is higher in the city with high economic level or neighbor to the coastal.Most cities prefer smart tourism which has a better economic pull effect.The tendency of cities to typical smart applications shows positive spatial autocorrelation characteristics,but their agglomeration areas are different.Among all the typical smart applications,the spatial autocorrelation characteristics of smart tourism and smart logistics are more obvious.The development of smart applications in different cities demonstrate three patterns,which are balanced,biased and short-board.Based on the analysis results,this paper provides suggestions to the development of China' s smart city from four perspectives,which are government control,application development,urban strategy,and mechanism improvement.In the discussion of data and research methods,the analysis summarizes their advantages through comparison of similar studies.The WeChat article data used in this research is timeliest,multidimensional,and easier to obtain than other used data.The text mining technology used in this research has been improved in its completeness,accuracy and comprehensiveness.Last,this paper summarizes the research methods and application scenario of the WeChat article data in urban studies from the aspect of time,space and contents.It also proposes a research framework to analyze policy contents based on text mining,which contains three key steps:keyword extraction,data preprocessing and text mining.The framework may provide guidance for future research and policy making.