{"updated":"2025-01-19T21:29:32.568320+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00199966","sets":["6164:6165:6210:9955"]},"path":["9955"],"owner":"44499","recid":"199966","title":["難しい詰めガイスター問題の生成法"],"pubdate":{"attribute_name":"公開日","attribute_value":"2019-11-01"},"_buckets":{"deposit":"cd8c5f96-b885-49ce-89ba-e704da42a4e6"},"_deposit":{"id":"199966","pid":{"type":"depid","value":"199966","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"難しい詰めガイスター問題の生成法","author_link":["485475","485478","485474","485477","485473","485479","485476","485480"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"難しい詰めガイスター問題の生成法"},{"subitem_title":"Generation of Difficult Geister Puzzle Instances","subitem_title_language":"en"}]},"item_type_id":"18","publish_date":"2019-11-01","item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_18_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"北陸先端科学技術大学院大学"},{"subitem_text_value":"松江工業高等専門学校"}]},"item_18_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Japan Advanced Institute of Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"National Institute of Technology, Matsue College","subitem_text_language":"en"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/199966/files/IPSJ-GPWS2019003.pdf","label":"IPSJ-GPWS2019003.pdf"},"date":[{"dateType":"Available","dateValue":"2019-11-01"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-GPWS2019003.pdf","filesize":[{"value":"1.8 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"5"},{"tax":["include_tax"],"price":"0","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"18"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"70c4428f-1c23-4726-aa67-5cb9614eefd5","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2019 by the Information Processing Society of Japan"}]},"item_18_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"石井, 岳史"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"川上, 直人"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"橋本, 剛"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"池田, 心"}],"nameIdentifiers":[{}]}]},"item_18_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Takefumi, Ishii","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Naoto, Kawakami","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Tsuyoshi, Hashimoto","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kokolo, Ikeda","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_5794","resourcetype":"conference paper"}]},"item_18_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"ボードゲーム『ガイスター』は6×6 のボード上で青赤2 種8 つの駒を交互に動かし，「脱出」「青駒全取り」「赤駒全取られ」のいずれかを狙う，互いの駒色がわからない2 人用不完全情報ゲームである．著者らはガイスターにおけるコンテンツとして詰めガイスター問題を提案したが，生成アルゴリズムの要因から 11 手詰めまでの問題しか生成できず，さらに問題の質を評価することができなかった．そこで本稿は，生成アルゴリズムにおける必勝手探索の探索法に Df-pn を用いることで大幅に探索速度を改善し，19 手詰め問題を得ることに成功した．それに加え，元の問題から手を戻すことで新たな問題を生成する逆順生成法を用いることで，狙った手数の問題の生成を可能とした．さらに，被験者実験を行い生成した問題の面白さと難しさについてアンケートを取り，教師あり学習を行うことで特徴量から面白さと難しさの推定を行った．推定誤差は5 段階評価の 0.5～0.6 程度で，ある程度の問題選別が可能であることを示した．","subitem_description_type":"Other"}]},"item_18_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"\"Geister\" is a two-player, zero-sum, deterministic but imperfect information board game. Each player plays using 4 blue and 4 red pieces, and the colors are hidden from the opponent player. \"Geister puzzle\" is a miniature problem of Geister as chess mating problem is to chess, where there is a way the player can win if no mistake was made. In the previous paper, we proposed a way to generate Geister puzzle instances. But the employed method was slow, so the maximum number of moves to win an instance was only 11. In this paper, we improved the generation method by using df-pn, and the maximum number of winning moves was increased to 19. In addition, we proposed a reverse generation method to improve the generation efficiency. Further, we tried supervised learning for making a prediction model of interestingness/difficulty of instances. The training data were collected from experiments using human subjects. We successfully trained models which can predict \ninterestingness/difficulty, where the root mean squared errors were around 0.5-0.6. ","subitem_description_type":"Other"}]},"item_18_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"19","bibliographic_titles":[{"bibliographic_title":"ゲームプログラミングワークショップ2019論文集"}],"bibliographicPageStart":"12","bibliographicIssueDates":{"bibliographicIssueDate":"2019-11-01","bibliographicIssueDateType":"Issued"},"bibliographicVolumeNumber":"2019"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:03:41.318138+00:00","id":199966,"links":{}}