TEMPORAL REASONING SYSTEM FOR THE DIGITAL THEATER LIBRARY

摘要

Each time actois appear on stage their performance is lost in the procession of temporality. However, scripts remain valuable research materials for digital library and theatrical analysis. Unfortunately in many cases, there is a lack of data about dates, performances, etc. The purpose of our research is to analyze the temporal structure of theatrical materials and implement a computer system to integrate this information. This paper focuses on Shinpa, often translated as "New School Theater" refers to a theatrical genre born in Meiji Japan (1868-1912). A Shinpa script was customarily rewritten each time as it was performed on stage, depending on the performing actors, the size of the theater, and so on, Thus, multiple acting scripts exist for a single drama, but they usually carry no information as to the printing or production dates. Although information about publication is necessary, theater researchers cannot depend solely on publication data such as production date, acting configuration and so on. In addition, within the current archival system there is inefficiency and waste In order to improve this situation, we implemented a temporal deduction system by using rule-based reasoning in artificial intelligence enabling us to automatically identify production date.