Please use this identifier to cite or link to this item:
|Title:||Analysis of Japanese folktales for the purpose of story generation||Authors:||Kato, T.
|Issue Date:||2008||Citation:||Kato, T.,Wama, T.,Miyazaki, K.,Nakatsu, R. (2008). Analysis of Japanese folktales for the purpose of story generation. Proceedings - 2008 the 1st IEEE International Conference on Ubi-Media Computing and Workshops, U-Media2008 : 447-452. ScholarBank@NUS Repository. https://doi.org/10.1109/UMEDIA.2008.4570933||Abstract:||We are developing an interactive folktale system in which users can enjoy both the generation of Japanese folktales as well as interaction with the system. In order to generate any kind of Japanese folktale, the system must give each character appearing in the folktales the ability to achieve a sufficient number of motions. To determine the number of motions that need to be prepared for the system, we carried out an analysis of the verbs that appear in Japanese folktales. Based on the analysis of 50 representative Japanese folktales, we found that the total number of verbs used is more than 2,000. By deleting the overlap of verbs in different folktales, while giving consideration to their meanings in actual usage, the number decreased to about 900. In addition, by restricting the verbs to those that can be expressed by computer graphics, the total number of verbs further decreased to about 100. Then we carried out experiments on generating various scenes of Japanese folktales by using several motion sets, each of which is a subset of the 100 motions corresponding to the reduced verbs. Finally, we evaluated the least number of animations needed to generate animations of sufficient quality. © 2008 IEEE.||Source Title:||Proceedings - 2008 the 1st IEEE International Conference on Ubi-Media Computing and Workshops, U-Media2008||URI:||http://scholarbank.nus.edu.sg/handle/10635/69394||ISBN:||9781424418657||DOI:||10.1109/UMEDIA.2008.4570933|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Jul 15, 2021
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.