{"created":"2025-01-19T01:13:12.864177+00:00","updated":"2025-01-19T17:34:13.051866+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00212218","sets":["1164:3500:10466:10598"]},"path":["10598"],"owner":"44499","recid":"212218","title":["Web上の画像から意味のある式を抽出するための数式フィルタの作成"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-07-23"},"_buckets":{"deposit":"0f738878-b55a-42a4-ac2b-56764ac7c0d4"},"_deposit":{"id":"212218","pid":{"type":"depid","value":"212218","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Web上の画像から意味のある式を抽出するための数式フィルタの作成","author_link":["540933","540932"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Web上の画像から意味のある式を抽出するための数式フィルタの作成"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"テキストの処理と理解","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2021-07-23","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"大阪市立大学"},{"subitem_text_value":"大阪市立大学"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Osaka City University","subitem_text_language":"en"},{"subitem_text_value":"Osaka City University","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"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/212218/files/IPSJ-IFAT21143003.pdf","label":"IPSJ-IFAT21143003.pdf"},"date":[{"dateType":"Available","dateValue":"2023-07-23"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-IFAT21143003.pdf","filesize":[{"value":"445.0 kB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"39"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"7380a12f-44f5-4fb5-8848-87a719a742f4","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"山田, 奉子"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"村上, 晴美"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10114171","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8884","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"数学情報検索（MIR）に関する研究は多数あるが，MathML や TeX で表された数式を解析するものがほとんどである．一方で，Web 上には数式画像を含む文書が多数存在するが，数式画像に関する研究は少ない．本研究は，Web 上の HTML 文書から，fh(x) = 1/nhΣni = 1K(x-xi/h) のような「コンパクトに表現された単独で意味のある数式」を抽出するものである．文書中の画像を教師あり機械学習の手法で，意味のある数式とそうでないものに二値分類する．深層学習以外の機械学習の手法と畳み込みネットワーク（CNN）の手法を比較した結果，VGG16 を用いた転移学習とファインチューニングを行った CNN の手法が一番良く，マシューズ相関係数（MCC）が 0.771，F 値が 0.814 であった．さらに応用事例として，数学用語をクエリとして Web 検索を行い，その用語と関連する数式を上位 10 件提示するという先行研究に対して，先行研究で使用した SVM による分類器を本研究で最も性能が良い CNN に置き換えたところ，F 値・MAP・MRR のいずれも先行研究を上回った．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告情報基礎とアクセス技術（IFAT）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2021-07-23","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"3","bibliographicVolumeNumber":"2021-IFAT-143"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":212218,"links":{}}