{"links":{},"id":2009652,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:02009652","sets":["1164:4088:1771221559804:1777431367596"]},"path":["1777431367596"],"owner":"80578","recid":"2009652","title":["ネットワーク運用の自動化を実現するAIを支えるMLOpsの取組紹介"],"pubdate":{"attribute_name":"PubDate","attribute_value":"2026-05-21"},"_buckets":{"deposit":"95a73e09-8505-48a6-98c8-f9f89d8d4d47"},"_deposit":{"id":"2009652","pid":{"type":"depid","value":"2009652","revision_id":0},"owners":[80578],"status":"published","created_by":80578},"item_title":"ネットワーク運用の自動化を実現するAIを支えるMLOpsの取組紹介","author_link":[],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"ネットワーク運用の自動化を実現するAIを支えるMLOpsの取組紹介","subitem_title_language":"ja"},{"subitem_title":"An Introduction to MLOps Supporting AI for Network Automation","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"ICM","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2026-05-21","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"NTT株式会社NTTネットワークイノベーションセンタ"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"NTT Network Innovation Center, NTT, Inc.","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/2009652/files/IPSJ-IOT26073007.pdf","label":"IPSJ-IOT26073007.pdf"},"date":[{"dateType":"Available","dateValue":"9999-01-01"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-IOT26073007.pdf","filesize":[{"value":"541.9 KB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"43"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"166b8178-1ba2-40f3-aef2-cd48c83b6185","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2026 by the Institute of Electronics, Information and Communication Engineers This SIG report is only available to those in membership of the SIG."}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"金森,圭太"}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Keita Kanamori","creatorNameLang":"en"}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA12326962","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":"2188-8787","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近年，ネットワーク(NW)の大規模化・複雑化に伴い，運用業務の効率化および高度化を目的としてAI（NW-AI）の活用が進んでいる．しかし，NWの網構成やトラフィック特性，運用ルールといった環境条件は時間とともに変化するため，導入時に高精度であったNW-AIも，環境変化に追従できなければ精度が徐々に劣化する。劣化したNW-AIを継続利用した場合，誤推定に基づく運用対応や障害対応の長期化，さらには計算資源および人的稼働の増加による運用コストの増大を招く恐れがある．NWの安定的な運用のために，NW-AIを継続的に運用・改善していくことが必要である．我々は，アラームを事象単位で集約するアラームコリレーションAIをはじめとするNW-AIを対象に，環境変化に適応し継続的に運用するためのシステム、MLOps（Machine Learning Operations）について研究してきた．本発表では，NWに特化したMLOpsの構成要素であるラベリング，モニタリング等の要素技術に関する技術課題について紹介する．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"In recent years, as networks (NW) have become increasingly large -scale and complex, the use of artificial intelligence for network operations (NW-AI) has been advancing in order to improve the efficiency and sophistication of operational tasks. On the other hand, environmental conditions such as network topology, traffic characteristics, and operational rules change over time. As a result, even NW-AI that achieved high accuracy at the time of deployment may gradually degrade in performance if it cannot adequately adapt to such environmental changes. Continued use of degraded NW-AI may lead to incorrect operational decisions and prolonged fault recovery processes. In addition, unnecessary processing may increase operational costs. Therefore, to ensure stable network operation, it is essential to contin uously operate and improve NW-AI. We have been conducting research on MLOps (Machine Learning Operations) as a system for continuously operating NW-AI while adapting to environmental changes, focusing on various NW-AI technologies such as alarm correlation method. In this presentation, we introduce technical challenges identified through the examination of an MLOps for operating NW-AI, specifically focusing on labeling, monitoring.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"1","bibliographic_titles":[{"bibliographic_title":"研究報告インターネットと運用技術（IOT）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2026-05-21","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"7","bibliographicVolumeNumber":"2026-IOT-73"}]},"relation_version_is_last":true,"weko_creator_id":"80578"},"created":"2026-05-14T07:16:32.930761+00:00","updated":"2026-05-14T07:16:39.138575+00:00"}