| The development and widespread use of internet technology and social media have significantly changed the ways in which the government and citizens communicate and interact.Government-citizen interaction platforms have become increasingly important channels for communication between the two parties,especially in light of the COVID-19 pandemic.However,extracting valuable information from massive government-citizen interaction data using modern computer technology remains a major challenge for government work.This information is crucial for accurately and timely responding to public demands.This study selects the Leader Message Board of People’s Daily Online as the research object.LDA topic modeling and sentiment analysis techniques are used to identify the themes and sentiment tendencies in the message data of 27 provinces and4 municipalities during the COVID-19 pandemic.Comparative analysis is employed to explore the similarities and differences in public demands across different provinces and cities.The goal is to provide reference for targeted policy adjustment and provision of public services at all levels of government,and to further enhance citizens’ participation in political decision-making and the democratic decision-making level of the government.In terms of identifying public demands,we classified the results of LDA topic model into 13 topic types according to content features,and analyzed the temporal evolution of topic strength for different demands.It was found that the topics of urban construction and housing were highly concerned by the public and showed a highly consistent evolution trend over time.The topics of water supply and heating,wage arrears,and education exhibited cyclic changes on an annual basis.The topic of epidemic prevention and control received high attention from the public during the early stage of the COVID-19 pandemic,but remained at a low level for over two years.Finally,combining the topic types with the identification of public sentiment inclination,we found that the topics related to daily life such as water supply and heating,community management,and consumer disputes had a relatively high proportion of negative comments.In terms of identifying inter-provincial differences in social demands,this study visualized the distribution of social demand texts from 31 provinces and municipalities using a percentage stacked bar chart.It was found that urban construction and housing were the most concentrated areas of social demand in most provinces,while the sum of the theme intensity of social security and epidemic prevention and control accounted for a relatively small proportion of messages across provinces.Subsequently,the top 5 themes with the highest proportion of topic strength in each province were calculated.It was found that urban construction,housing,education,community management,and living environment were common high-frequency issues of concern among all provinces.Among them,the housing issue had the highest proportion of topics in Hainan,the living environment issue had the highest proportion of topics in Guangdong,and the consumer dispute issue had the highest proportion of topics in Shanghai.This article provides certain policy suggestions for these issues.Finally,this study compared the distribution of sentiment tendencies among provinces and municipalities and found that Tibet and Xinjiang had the highest proportion of positive emotions.This indicates that China’s frontier policies have played an important role in promoting local economic and social development,and improving the happiness and sense of belonging of local residents.In addition,due to significant regional differences among provinces,public demands have complex and diverse characteristics.Therefore,this article provides specific suggestions on strengthening policy information disclosure,promoting resource integration of online petition platforms,establishing multi-departmental collaborative mechanisms,and increasing publicity efforts for government-citizen interaction platforms. |
PDF Full Text Request