|
パテント・インテグレーションは2026年9月10日、生成AIによる特許調査サービス「サマリア」が、2024年度特許検索競技大会アドバンストコースの過去問検証において、電気・機械・化学/医薬の3分野すべてで「再現率100%(上位30件)」を達成したと発表しました。特許調査への生成AI活用が進むなか、検索式の語彙問題という課題を、「類語シソーラス拡張」機能で対応した成果です。(「類語シソーラス拡張」機能とは、過去の拒絶理由通知書10万件以上から、審査官が本願発明と引用文献それぞれの用語・概念を対応するものと認定した事例を抽出し、調査時の類語辞書として整備した機能。) この発表について、ClaudeとChatGPTに深掘り調査・分析させましたので、ご参照ください。なお、生成AIによる調査・分析結果は、公開された情報だけに基づくものであり、必ずしも実情を示したものではなく、誤った情報を含む可能性があることにご留意のうえ、ご参照ください。 生成AIによる特許調査サービス「サマリア」、特許検索競技大会の過去問検証で全3分野の再現率100%を達成 https://note.com/patent_i/n/nc5d61858bb7d Summaria Achieves “100% Recall” on Past Patent Search Competition Questions On September 10, 2026, Patent Integration announced that its generative AI-powered patent search service, Summaria, had achieved “100% recall (top 30 results)” across all three technical fields—electrical engineering, mechanical engineering, and chemistry/pharmaceuticals—in an evaluation using past questions from the Advanced Course of the 2024 Patent Search Competition. As the use of generative AI in patent searches continues to expand, this achievement demonstrates how the service’s “Synonym Thesaurus Expansion” feature addresses vocabulary-related challenges in formulating search queries. (The “Synonym Thesaurus Expansion” feature draws on more than 100,000 past notices of reasons for refusal to extract instances in which examiners identified correspondences between terms and concepts used in the inventions under examination and those in cited references. These correspondences are compiled into a synonym dictionary for use in patent searches.) I asked Claude and ChatGPT to conduct in-depth research and analysis of this announcement, and their findings are presented for your reference. Please note, however, that these AI-generated research and analysis results are based solely on publicly available information, may not necessarily reflect the actual circumstances, and may contain inaccuracies. Your browser does not support viewing this document. Click here to download the document. Your browser does not support viewing this document. Click here to download the document. Your browser does not support viewing this document. Click here to download the document. Your browser does not support viewing this document. Click here to download the document. Your browser does not support viewing this document. Click here to download the document. Your browser does not support viewing this document. Click here to download the document.
0 Comments
Leave a Reply. |
著者萬秀憲 アーカイブ
April 2026
カテゴリー |
RSS Feed