Spatial-temporal Change Of Citizens' Participation In Urban Governance For SDGs

The United Nations Sustainable Development Goals(SDG s)have become the consensus of national development.In the sustainable development goals of cities(SDG 11),it is an important indicator to build sustainable cities to strengthen inclusive sustainable urban construction and cultivate participatory,comprehensive and sustainable community planning and urban management capabilities(SDG 11.3.2).As a platform for the government to implement the interaction between the government and the citizens,the government website has become a new bridge for the citizens to participate in urban governance.Its participation degree shows a city's co-governance ability and is an important reference index for evaluating the sustainability of a city.The booming big data technology and deep learning methods provide a new perspective for the study of urban sustainable development.With the improvement of citizens' democratic consciousness and the continuous development of democratic administration,the political and civil network interaction platform,as a kind of social service platform,has entered into citizens' life.Its massive information,unique user interaction mode and virtual social space attract a large number of users to participate.The government's effective use of the network interaction platform is conducive to promoting communication with citizens and understanding the city People's concerns are conducive to the sustainable development of cities.In this paper,aiming at SDG 11.3.2,Taking Suzhou as an example,through in-depth learning and data mining of about 6 million text data of citizens' participation in urban governance in Suzhou "Hanshanwenzhong" forum from 2012 to 2019,a quantitative model of SDG 11.3.2 in Suzhou is designed,and the spatial-temporal pattern evolution law of SDG 11.3.2 in Suzhou is studied on different spatial-temporal scales.The main processing flow includes: 1)construct Suzhou geographic entity recognition and annotation data set,subject classification and annotation data set,and emotion analysis data set.Based on the pre training language model Bert,the text data of Suzhou ‘Hanshanwenzhong' forum are deeply learned and geographical entity recognition,subject classification and emotion analysis are carried out.2)Based on the results of geographical entity recognition,each topic is related to different spatial scales and specific spatial locations in Suzhou.3)Based on the results of text emotion analysis,the emotional satisfaction of residents on different topics was calculated.4)From different time and space scales,this paper discusses the hot issues and their spatial distribution of citizens' participation in urban governance,as well as their emotional tendency and changing rules.5)Based on the data of spatial location and emotional results discussed in the forum,the SDG 11.3.2 quantitative model with different spatial and temporal scales in Suzhou was designed,and its citizen participation was evaluated.The main conclusions are as follows:(1)From the content of citizens' participation in the discussion,it covers a wide range of topics,including all aspects of daily life,and each question has received the timely attention of relevant government personnel.Nearly half of the questions were answered more specifically by the official department within one week.(2)From the perspective of user interaction content,the topic content mainly focuses on issues related to living quarters and daily travel,and 80% of complaints information,with negative emotional tendency.And through empirical research,the main factors that affect people's complaints are the government's new policies and the inconvenience caused by the implementation of urban planning and construction,among which the most obvious impact is caused by the construction and transformation of residential property and roads.(3)From the perspective of user interaction time,under different time scales,the heat of public participation in urban governance shows different rules,and it will be significantly reduced in holidays,reducing to one-half of the original.(4)From the perspective of spatial distribution of geographical entities involved,the interactive content of geographical entities involved is concentrated in Wuzhong District(19.74%),Gusu district(18.57%)and Industrial Park(18.02%).The above conclusions can provide a reference for the government to accurately grasp the participation of urban governance in the big data environment and formulate corresponding policy guidance strategies.