Correlation Between Public Sentiment And Built Environment In Shenzhen Based On Weibo Data

In the rapid development of China's cities,it has changed from a sprawling development to an internal renewal of built-up areas,and has paid more attention to the quality of urban life.The construction of a "happy city" proposed by China needs to pay more attention to human perception and experience.In the existing research,more attention is paid to t he built environment and behavioral activities and physical health,and few people pay attention to public emotions and mental health.The ever-maturing mobile network and big data analysis technologies provide data for obtaining public subjective mood and attitude data Strong support makes it possible to study emotional maps.Considering the multi-classification of emotions and the final accuracy rate,the study determined to use machine learning and other technical methods to perform sentiment analysis on Shenzhen's Weibo text data.In order to fully describe the characteristics of emotions,emotions are divided into three dimensions of intensity,polarity,and motivation and six types of emotions,which are happy,like,expect,sad,disgusted,and angry,respectively.Based on this type of emotional map,the spatial and temporal distribution characteristics of public emotions are described.The analysis shows that the intensity,polarity,and motivation of public emotions during holidays are higher than th at of working days,and high-intensity and positive emotions are gathered on a large scale in the central urban area.Low-intensity and negative emotions are mainly concentrated in the industrial parks in the outer urban areas.District has the greatest emotional motivation.Based on the existing related research and the current situation of emotion distribution,16 built environment elements were selected from the five aspects of land use,space form,road traffic,green space and open space,and daily life service facilities.At the same time,emotions in three different emotional dimensions were selected.The built environment characteristics of the area are analyzed.Through SPSS,correlation analysis of built-up environmental element data with different dimensions and types of emotion was performed.The study has shown that emotional polarity is significantly related to most built-up environmental factors.Emotional intensity is negatively correlated with industrial land and has the greatest correlation.Emotional polarity is positively correlated with distance from the nearest subway station and has the greatest correlation.Emotional motivation is related to building density.It is negatively correlated and has the greatest correlation.The two types of emotional types,"Like" and "Sad" are closely related to the built environment and are significantly related to most of the built environment elements.Based on the correlation between public sentiment and the built environment,reflections and suggestions for planning are proposed at a macro level.