Gaining deep insights of the social Web content is a challenging Big Data analytics problem, especially when dealing with social posts of high volume and arrival rate consisting of high variable topics. Detecting and tracking the topics that the users discuss in popular microblogging applications like Twitter and studying the evolution of each topic reveals crowd interests and intelligence. The evolution summarizes the changes that are occurring on the topics over a given time horizon inside the evolving data stream. For instance, some topics may disappear at some point in time due to lack of users' interest, while others are retained over time adopting either a stable or an evolving behaviour. The analysis and storage of such dynamic and massive content with spatio-thematic properties poses new challenges for research.