Political polarization has become an increasing problem in the US and has caused concern also in many other countries. The lecture presents different ways of conceptualizing polarization with a special focus on structural polarization. First, polarization can refer to issue or ideological polarization where political views move further away from each other. Second, affective polarization means that negative partisanship increases as those parties that one does not support are seen in an increasingly negative light. Lastly, structural polarization is a network-based way of analyzing in-group and out-group interactions with the insight that the less interaction there is between groups, the more polarized they are. The lecture shows how the structural conceptualization is especially suited for studying polarization with big data from social media platforms. The lecture also suggests that the structural conceptualization of polarization makes it possible to study both the ideological and affective aspects of polarization on social media. The lecture reports results from a longitudinal study of the polarization of five topics (immigration, climate change etc.) on Finnish social media on the platform X in 2015–2023 (11.5 million tweets). The results show that social media users can be categorized into ideological groups based on their retweeting behavior. In addition, tracking the polarization and activity of these ideological groups shows which groups are responsible for the polarization of the discussion, how polarization develops in time, and what are the causes driving increases in polarization. The lecture concludes by presenting ideas on how to study affective polarization from a network perspective with social media data.