Social media and its analogous applications allow millions of users to express and spread their opinions about a topic and show their attitudes by liking or disliking content. The amount of data accumulating on social media is known as Big Social Data which can be analysed to provide valuable insight in many application contexts. Measuring public opinion during events like elections is one of the most promising BSD applications either for predicting results or analysing online debate during electoral campaigns. In this paper, the analysis of political debate on social media during the 2020 regional election in Emilia Romagna (Italy) is presented. Candidates’ programs have been analysed in order to extract topics within the official action plans. Leveraging Big Social Data, the interactions of Italian citizens on Twitter candidates’ pages have been monitored and analysed, in order to understand how action plans within candidates’ programs has been perceived by citizens through the lens of social media; which are the most relevant topics emerged from online debate in the social networks; which is the sentiment emerged in the overall debate; how the online debate have polarized on emerged topics and how certain communities of users have contributed to the information diffusion. Results have shown a mismatch between the action plans in the candidates programs and the topics emerged online debate, along with a unidirectional communication approach followed by the three candidates that has resulted in the polarization of the debate on some of the topics emerged during the campaign.