Designing serious games with linguistic resources

Language or word based casual games are becoming increasingly popular, but many of the game-play mechanics rely on spelling and word searching. This sub-genre of casual games is also targeted by serious games developers, but in their current 'spelling-bee' form they are not ideal tools for pedagogical assessment. Incorporating meta-data about linguistic constructs and their relationships would enable creative uses of existing corpora in games. Unfortunately, an impediment to integrating these linguistic resources into games is the laborious task of compiling and condensing the available data into a suitable form. This research into creative serious games has developed a word matching game which utilizes multiple diverse corpora to create the game-play data. The game itself is incorporated as an interface to an Intelligent Tutoring System framework. The design of the game and the approach to data generation is described.