Reading Between the Lines: Using Plot Graphs to Draw Inferences from Stories

Intelligent agents designed to interact with humans need to be able to understand human narratives. Past attempts at creating story understanding systems are either computationally expensive or require a vast amount of hand-authored information to function. To combat these difficulties, we propose and evaluate a new story understanding system using plot graphs, which can be learned from crowdsourced data. Our system is able to generate story inferences much quicker than the baseline alternative without significant loss of accuracy.