Gathering Background Knowledge for Story Understanding through Crowdsourcing

Successfully comprehending stories involves integration of the story information with the reader’s own background knowledge. A prerequisite, then, of building automated story understanding systems is the availability of such background knowledge. We take the approach that knowledge appropriate for story understanding can be gathered by sourcing the task to the crowd. Our methodology centers on breaking this task into a sequence of more specific tasks, so that human participants not only identify relevant knowledge, but also convert it into a machine-readable form, generalize it, and evaluate its appropriateness. These individual tasks are presented to human participants as missions in an online game, oering them, in this manner, an incentive for their participation. We report on an initial deployment of the game, and discuss our ongoing work for integrating the knowledge gathering task into a full-fledged story understanding engine. 1998 ACM Subject Classification I.2.4 Knowledge Representation Formalisms and Methods