In this paper we describe Answery, a rule-based system that allows authors to specify game characters' background stories in natural language. The system parses these background stories, applies transfor- mation rules to turn them into semantic content, and generates dialogue during gameplay by posing it as a question-answering problem. By the means of simple categorization combined with rule inference engine, our system can generate answers eciently. Our initial pilot study shows that this approach is promising.
Gamnlp 12. Proceedings of the First Workshop on Games and Nlp, Japtal 2012: In Connection With the 8th International Conference on Natural Language Processing, 2013
story; comprehension; dialogue
Main Research Area:
CfP: 1st Workshop on Games and Natural Language Processing, 2013