{"id":2000033,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:02000033","sets":["581:11839:11841"]},"path":["11841"],"owner":"80578","recid":"2000033","title":["施策効果の高い顧客グループの発見を目的とした機械学習に基づく実験計画手法"],"pubdate":{"attribute_name":"PubDate","attribute_value":"2025-02-15"},"_buckets":{"deposit":"5828fd7d-9395-4f28-8bbb-d1a7f225d153"},"_deposit":{"id":"2000033","pid":{"type":"depid","value":"2000033","revision_id":0},"owner":"80578","owners":[80578],"status":"published","created_by":80578},"item_title":"施策効果の高い顧客グループの発見を目的とした機械学習に基づく実験計画手法","author_link":[],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"施策効果の高い顧客グループの発見を目的とした機械学習に基づく実験計画手法","subitem_title_language":"ja"},{"subitem_title":"Experimental Design Based on Machine Learning for Finding Customer Groups with High Measure Effects","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"[一般論文] 実験計画，機械学習，効果検証，施策効果，Eコマースマーケティング","subitem_subject_scheme":"Other"}]},"item_type_id":"2","publish_date":"2025-02-15","item_2_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"早稲田大学"},{"subitem_text_value":"早稲田大学"},{"subitem_text_value":"株式会社ZOZO"},{"subitem_text_value":"早稲田大学"}]},"item_2_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Waseda University","subitem_text_language":"en"},{"subitem_text_value":"Waseda University","subitem_text_language":"en"},{"subitem_text_value":"ZOZO, Inc.","subitem_text_language":"en"},{"subitem_text_value":"Waseda University","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"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/2000033/files/IPSJ-JNL6602032.pdf","label":"IPSJ-JNL6602032.pdf"},"date":[{"dateType":"Available","dateValue":"2027-02-15"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-JNL6602032.pdf","filesize":[{"value":"1.1 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"8"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"8b52d64c-a858-4f4c-ab7d-acd05ec95770","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2025 by the Information Processing Society of Japan"}]},"item_2_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"中村,友香"}]},{"creatorNames":[{"creatorName":"山極,綾子"}]},{"creatorNames":[{"creatorName":"佐々木,北都"}]},{"creatorNames":[{"creatorName":"後藤,正幸"}]}]},"item_2_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Yuka Nakamura","creatorNameLang":"en"}]},{"creatorNames":[{"creatorName":"Ayako Yamagiwa","creatorNameLang":"en"}]},{"creatorNames":[{"creatorName":"Hokuto Sasaki","creatorNameLang":"en"}]},{"creatorNames":[{"creatorName":"Masayuki Goto","creatorNameLang":"en"}]}]},"item_2_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN00116647","subitem_source_identifier_type":"NCID"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_6501","resourcetype":"journal article"}]},"item_2_publisher_15":{"attribute_name":"公開者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"item_2_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1882-7764","subitem_source_identifier_type":"ISSN"}]},"item_2_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"ECサイト上でビジネスを行う多くの企業にとって，クーポン配布などのマーケティング施策は一般的になっており，より効率的に行う方法に関する研究がさかんに行われている．これらの施策はユーザの離反を防ぐとともに定着化を図ることを目的に行われるが，全ユーザに一様に施策を実施することは必ずしも得策ではない．施策による影響の大小は個人差があるため，効率的かつ効果的に施策を講じるためには施策によってLife Time Value（LTV）の向上が見込まれるような施策効果が高い顧客群を特定することが重要である．本研究では，ファッション通販サイトZOZOTOWN内で古着を販売するZOZOUSEDを対象事例とし，ポイント付与などの施策を講じる対象ユーザの選定に関する問題を考える．これまでZOZOUSEDでは，独自のロジックに基づき担当者が自らの知見から施策対象のユーザグループを選定していたが，大量の未開拓のユーザ群の中に高い効果が望めるユーザ群が存在する可能性がある．しかし，大多数の未開拓ユーザの中から効果の高いユーザを発見するための施策実験を行おうとした場合，実験的に行う施策対象者数にはコスト面での上限があり，加えて施策対象者に施策効果が低いユーザを多く含んでしまうことによる利益減少リスクもともなう．そこで本研究では，未開拓のユーザ群の中から高い効果が望めるユーザ群を特定するための定量的指標に基づいた実験計画手法を提案する．具体的には，予測購買単価と予測利用回数の切り口でユーザをセグメント化し，各セグメントにおける施策効果の平均と分散を基に施策対象とするセグメントを決定することで，より効率的な高効果ユーザ群の特定を目指した分析アプローチを構築する．最後に，ZOZOUSEDで行った検証実験を基に提案手法の有用性を示す．","subitem_description_type":"Other"}]},"item_2_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"For many companies doing business on e-commerce sites, marketing measures such as coupon distribution have become common, and research on more efficient methods has been active. However, it is not necessarily advisable to implement measures such as coupon distribution uniformly to all users. Since the impact of a measure differs depending on the individual, it is important to identify a group of customers for whom the measure is highly effective in order to implement the measure efficiently and effectively. In this study, we consider the problem of target user selection for the point award policy as a promotion measure, focusing on ZOZOUSED, which sells used clothes on the fashion shopping site ZOZOTOWN, as the target case study. Currently, as with most e-commerce sites, most of the target users are in the reserve group for defection. There is a possibility that there are unexplored user groups that can be highly effective. Therefore, this study considers the design of a measure experiment to discover a group of users with high measure effectiveness among such a large number of unexplored users. However, in conducting a verification experiment, there is a risk of including many target users with low effectiveness and of restricting the number of target users. In particular, there is a possibility that an unexploited user group exist among the non-respondents of the measures that could be effective in improving its' effect. Therefore, this study proposes an experimental design method based on quantitative indicators estimated by a machine learning model for identifying expected high-effective user groups from among undeveloped user groups.","subitem_description_type":"Other"}]},"item_2_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"439","bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌"}],"bibliographicPageStart":"426","bibliographicIssueDates":{"bibliographicIssueDate":"2025-02-15","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"2","bibliographicVolumeNumber":"66"}]},"relation_version_is_last":true,"item_2_identifier_registration":{"attribute_name":"ID登録","attribute_value_mlt":[{"subitem_identifier_reg_text":"10.20729/0002000033","subitem_identifier_reg_type":"JaLC"}]},"weko_creator_id":"80578"},"updated":"2025-02-18T02:15:14.490042+00:00","created":"2025-02-12T00:48:28.011968+00:00","links":{}}