グリムベルゲン　ライエル   Ｇｒｉｍｂｅｒｇｅｎ　Ｒｅijer    所属   コンピュータサイエンス学部 コンピュータサイエンス学科    職種   教授
言語種別	英語
発行・発表の年月	2005/09
形態種別	国際会議論文
査読	査読あり
標題	A Cognitive Science Approach to Shogi Playing Processes - Some Results on Next Move Test Experiments -
執筆形態	共著
掲載誌名	In: Van den Herik, H., Hsu, S-C., Hsu, T. and Donkers, H.H.L.M. (Eds.), Advances in Computer Games, 11th International Conference, ACG2005
出版社・発行元	Taipei, Taiwan
巻・号・頁	LNCS 4250,pp.140-154
著者・共著者	Ito, T., Matsubara, H. and Grimbergen, R.
概要	In the past, there have been numerous studies into the cognitive processes involved in human problem solving. From the start, games and game theory have played an important role in the study of human problem solving behavior. In chess, several cognitive experiments have been performed and those experiments have shown that expert chess players can memorize positions very quickly and accurately. Chase and Simon introduced the concept of chunking to explain why expert chess players perform so well in memory tasks. Chunking is the process of dividing a chess position into smaller parts that have meaning. We performed similar experiments in Shogi with a set of next move problem positions, collecting verbal protocol data and eye movement data. Even though experiments in chess indicated that expert chess players searched as wide and deep as non-expert players, our experiments show that expert Shogi players search more moves, search deeper and search faster than non-expert players. Our experiments also show that expert Shogi players can not only memorize the patterns of the positions but also recognize move sequences before and after the position. The results suggest that other than the perceptual spatial chunks introduced in chess research, there are also chunks of meaningful move sequences. We call such chunks “temporal chunk”. Our research indicates that Shogi players become stronger by acquiring these temporal chunks.