Tatsuya Arai


2016

pdf
Using Synthetically Collected Scripts for Story Generation
Takashi Ogata | Tatsuya Arai | Jumpei Ono
Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: System Demonstrations

A script is a type of knowledge representation in artificial intelligence (AI). This paper presents two methods for synthetically using collected scripts for story generation. The first method recursively generates long sequences of events and the second creates script networks. Although related studies generally use one or more scripts for story generation, this research synthetically uses many scripts to flexibly generate a diverse narrative.