Narrative intelligence, or the ability to tell and understand stories, is a vital component of human cognition. Converging evidences suggest narrative intelligence is inherently knowledge intensive. The limited amount of knowledge available restricts most computational systems to micro-worlds and prevents large-scale application of narrative intelligence.
This research aims to address the knowledge bottleneck for computational narrative intelligence. Our technique acquires knowledge needed for narrative intelligence in a quick, economical, and user-friendly manner. The learned knowledge can be utilized to generate interactive fictions. The acquisition and application of the knowledge is bridged by a script-like knowledge representation that caters to both processes.