This paper describes a method and a toolset for capturing and analysing public opinions in social media in order to support citizen participation in urban development projects. This method was developed and is currently applied in the EU Horizon 2020 project “U_CODE” in order to inform professional urban planners about the public opinion and its different variations. In contrast to other projects our approach explicitly takes into account the dialogical nature of discourse in social media in order to capture the dynamics of online discourse and to get deeper insights into the public debate. We use a method called Discourse Structure Analysis to analyse the often complex message threads in social media. This method is able to detect dialogue sequences in order to find intensive discussions and calculate the amount of participation in relation to different topics. By analysing the reply patterns we derive the social network structures of the participants and find sub-groups and citizen experts. In combination with topic and sentiment analysis we can create structures which represent the connections between topics, sentiments and people and visualize these in different ways in order to support the cooperation between the public and the professional planners. Topic and sentiment analysis are very popular in the field of market research and arose as a helpful tool to gain useful insights from Social Media data. State of the art approaches usually share the method of assigning a numeric score between -1 and 1 to a target word referring to a product, a company or other developments to be evaluated. However, a single score which indicates whether sentiments are positive, negative or neutral, do not offer many insights for adjustment. Therefore, a combined approach of discourse, text and sentiment analysis for target based opinion retrieval is better suited to detect wishes, concerns, fear and similar emotions but also problems and ideas which are related to an urban planning project. The outcomes of the application are the following: Automatic capturing of online discourse across different social media channels. Mapping of public opinion structures, social networks and their interplay. Visualization of online debates in order to derive and summarize core argumentation structures. Feedback to all stakeholders of the discussion (participants, planners, decision makers) by adapted analysis results and visualizations.