With the advent of the new media era, the form of public participation in urban planning has become more diversified, and its mode and extent have undergone profound changes. Traditional ways in which public opinions are manually selected and summarized have failed to meet the needs of processing an ocean of public opinions. Thus, it is of urgent necessity to apply big data mining and analysis technology to obtaining,extracting and summarizing public opinions. Taking Sina Weibo, a main social media platform in China, as a source of public opinions, this study crawled 12 171 microblog data related to Shanghai Master Plan (2017-2035) (hereinafter referred to as ″Shanghai 2035″) during its planning. Natural language processing technology was employed to obtain public opinions, identify public attitudes and analyze public behaviors, thereby revealing the results and features of internet-based public participation in Shanghai 2035. To better coordinate the interests of various stakeholders and boost the level of planning in the long term,an integrated topical and emotional analysis was made to get a comprehensive picture of public demands of values and to track down the development of these issues. The public gave huge attention to Shanghai 2035 on Sina Weibo with respect to the various aspects of planning. The three most attention-getting topics were ″population″, ″global city″, and ″publicity of the draft″. Topics such as ″housing price″, ″construction land″, ″metropolitan area″, and ″transportation″ were also the focus of public attention. The results of the overall sentiment analysis show that the proportion of microblog data with negative and positive emotions amounted to 56.69% and 39.97%, respectively. Almost all the top eight topics were dominated by negative opinions, except for ″publicity of the draft″ to which a slightly larger number of people responded positively. In brief, most of the public on Sina Weibo held negative attitudes toward the topic, be it overall sentiment or topical sentiment. The spatial-temporal analysis of topic evolution suggests that public attention to hot topics is not continuous; rather, it is intermittent and explosive and the public in different districts are concerned about different topics. This phenomenon is closely related to socioeconomic situations and planning impacts, indicative of the spatial-temporal preference of planning as a public policy. In this paper, the spatial mobility of topics is further explored by taking two trending topics (i.e., ″total population″ and ″global city″) in all districts as examples. The result presents an obvious spatial imbalance in terms of topic mobility.The public in the main urban zone played a leading role in public participation in Shanghai 2035 in that topics were both initiated and ended in the main urban zone whereas people in suburban areas participated in discussions only during the heated discussion period. With regard to public behaviors, public participation on a social media platform is featured by ″the government posting microblogs and the general public commenting on and reposting microblogs″. The three types of Weibo subscribers — Blue V, Yellow V and general users — represent different stakeholders with different interests and demands, such as government agencies, experts in relevant fields and common users. They focused on different topics and expressed different emotions. Blue V and Yellow V users were concerned about more strategic topics with more positive attitudes, while general users paid more attention to topics closely related to their daily life and most of them held negative attitudes. The disparity between one’s current life situation and the planning horizon could be the primary cause for negative emotions among the general public.