EventGo! Exploring Event Dynamics from Social-Media Posts

Looking for local events and promotions is a common need for most people during travel or moving to a new city. Similarly, delivering event messages to the right people is also a challenge for small businesses that seek ways to promote their events. In addition to announcements on their websites, further advertisement on social networking sites is very common, e.g., Facebook fanpages or event posts, and event-based social networks such as meetup.com. The goal of this paper is to build an event search engine to fulfill the information need. We start with the event source from social networks and design event title and place name recognition models for event database construction. The framework enables the extraction of 1K events from 230K Facebook fanpages in Taiwan every day. With the construction of event database from social networks, we are able to explore the dynamics of each city in Taiwan and disclose statistics in the events extracted from Facebook fanpages and Facebook events, showing the change in the ad market.