With the advent of smartphones and auto-uploaders, user-generated content (e.g., tweets, photos, and videos) uploads on social media have become more numerous and asynchronous. Thus, it is difficult and time taking for users to manually search (detect) interesting events. It requires for social media companies to automatically detect events and subsequently recommend them to their users. An automatic event detection is also very useful in an efficient search and retrieval of user-generated content. Furthermore, since the number of users and events on event-based social networks (EBSN) is increasing rapidly, it is not feasible for users to manually find the personalized events of their interest. We would like to further explore events on EBSN such as Meetup for different multimedia analytics projects such as recommending events, groups, and friends to users. At MIDAS@IIITD, we would like to use Deep Neural Network (DNN) technologies due to their immense success to address these interesting problems. Our recent papers on event detection and summarization are published in top-tier conferences and journals such as Knowledge-Based Systems, ACM Multimedia, ACM ICMR, etc.