Interactive Narrative Generation Using Location and Genre Specific Context

In this paper, we present a novel method of creating interactive narratives. Location-based narrative sentences are mapped to their real-world location using GPS. A corpus of genre-based sentences is aggregated to train a neural network. New narratives are created by selecting a sentence from the network’s “Sentence Space” that corresponds to contextual information from the user’s physical location. Each narrative is explored using GPS location tracking and WebGL based augmented reality in mobile web browsers.