Just-in-time reconnaissance and assistance for video game streams and players

While video game play streaming and archiving through sites such as Twitch.tv is an increasingly relevant and commercially viable practice, limitations of many of those systems ultimately cost end-users the chance to view the most-relevant streams at the moment they need them. Furthermore, while SDKs for in-game event and metric gathering are becoming prevalent, they are not typically combined with real-time pattern matching capabilities of any sophistication and, therefore, will generally fail to allow `deep' matching between gamers and the game streams of highest relevance. This work in progress paper reports on a framework that will, when complete, fill major gaps in video game stream profiling, indexing, search, and usability (for players), and will be positioned as an offering for game developers and aggregators.