[3D1-OS-22a] OS-22 (1)

[3D1-OS-22a-03] Introducing an Extensive Knowledge Generation Mechanism into a Narrative Generation System

〇Kazui Fukuda¹, Jumpei Ono^2,3, Takashi Ogata³ (1. Faculty of Policy Studies, Iwate Prefectural University , 2. Vocational School of Digital Arts Sendai, 3. Faculty of Software and Information Science, Iwate Prefectural University)

Keywords:Narrative Generation System, Conecrptual Dictionary, Property Information Acquisition, Unchiku, Creativity

The authors have been developing an “integrated narrative generation system” that generates the stories, deep rhetorical structures (narrative discourses), and surface representations in a narrative. In this paper, for the integrated narrative generation system, the authors present a mechanism that can generate “unchiku”, the detailed and excessive knowledge contents regarding a specific object, theme, or topic. In particular, the authors store the attribute information about each noun concepts acquired automatically using the Japanese Wikipedia into the noun conceptual dictionary in the integrated narrative generation system. The proposed unchiku generation mechanism enables to generate unchiku information about various objects and topics through inserting the parts of acquired unchiku knowledge contents into various points in the story generated by the integrated narrative generation system. In this time, attribute information related to kabuki is collected from the Japanese Wikipedia and used in this proposed unchiku generation mechanism.