Revealing the Hidden Links in Content Networks: An Application to Event Discovery

摘要: Social networks have become the de facto online resource for people to share, comment on and be informed about events pertinent their interests livelihood, ranging from road traffic or an illness concerts earthquakes, economics politics. This has been driving force behind research endeavors that analyse such data. In this paper, we focus how Content Networks can help us identify effectively. incorporate both structural content-related information of a social network in unified way, at same time, bringing together two disparate lines research: graph-based content-based event discovery media. We model interactions types nodes, users content, introduce algorithm builds heterogeneous, dynamic graphs, addition revealing content links network's structure. By linking similar nodes tracking connected components over effectively different events. Our evaluation media streaming data suggests our approach outperforms state-of-the-art techniques, while showcasing significance hidden quality results.