An automatic rule creating method for Kansei data and its application to a font creating system

Hajime Hotta, Masafumi Hagiwara

研究成果: Conference contribution

4 被引用数 (Scopus)

抄録

In this paper, we propose a method for creating fuzzy rules of Kansei data automatically. This method consists of 3 steps: (1) Generation of pseudo data of Kansei data by a General Regression Neural Network; (2) Clustering the pseudo data by a Fuzzy ART; (3) Translating each cluster into a fuzzy rule and extracting important rules. In this experiment, we applied this method to "a Japanese font creating system reflecting user's Kansei ." From the result of the experiment, although we have used the same algorithm for drawing font outlines, the system employing our method can reflect Kansei better than the conventional one.

本文言語English
ホスト出版物のタイトルModeling Decisions for Artificial Intelligence - Second International Conference, MDAI 2005, Proceedings
出版社Springer Verlag
ページ421-430
ページ数10
ISBN(印刷版)3540278710, 9783540278719
DOI
出版ステータスPublished - 2005
イベント2nd International Conference on Modeling Decisions for Artificial Intelligence, MDAI 2005 - Tsukuba, Japan
継続期間: 2005 7月 252005 7月 27

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
3558 LNAI
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Other

Other2nd International Conference on Modeling Decisions for Artificial Intelligence, MDAI 2005
国/地域Japan
CityTsukuba
Period05/7/2505/7/27

ASJC Scopus subject areas

  • 理論的コンピュータサイエンス
  • コンピュータ サイエンス(全般)

フィンガープリント

「An automatic rule creating method for Kansei data and its application to a font creating system」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル