MultArtRec: A Multimodal Neural Topic Model for Integrating Image and Textual Features in Artwork Recommendation

• Item type: SIG Technical Reports(1)
• 公開日: 2023-09-14
• タイトル: MultArtRec: A Multimodal Neural Topic Model for Integrating Image and Textual Features in Artwork Recommendation
• タイトル / 言語: en
• タイトル: MultArtRec: A Multimodal Neural Topic Model for Integrating Image and Textual Features in Artwork Recommendation
• 言語: eng
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Graduate School of Information Science and Engineering, Ritsumeikan University
• 著者所属: College of Information Science and Engineering, Ritsumeikan University
• 著者所属: College of Information Science and Engineering, Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): Graduate School of Information Science and Engineering, Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): College of Information Science and Engineering, Ritsumeikan University
• 著者所属(英): en
• 著者所属(英): College of Information Science and Engineering, Ritsumeikan University
• 著者名: Jiayun, Wang; Akira, Maeda; Kyoji, Kawagoe
• 著者名(英): Jiayun, Wang; Akira, Maeda; Kyoji, Kawagoe
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Multimodal models have demonstrated remarkable success in the domains of image processing and natural language processing. Recently, their significance has also been acknowledged within recommendation systems. In many cases, the recommendation systems perform better when utilizing multimodal features to construct item embeddings, rather than utilizing individual text or image models. Consequently, research in this field has shifted its focus towards effectively combining multimodal features and accurately embedding items. Our study specifically concentrates on artwork recommendation. In artwork recommendation, the textual data such as titles and descriptions notably influence users' preferences. Our research approach involves constructing multimodal embeddings of artworks by integrating both images and titles as a fundamental step.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Multimodal models have demonstrated remarkable success in the domains of image processing and natural language processing. Recently, their significance has also been acknowledged within recommendation systems. In many cases, the recommendation systems perform better when utilizing multimodal features to construct item embeddings, rather than utilizing individual text or image models. Consequently, research in this field has shifted its focus towards effectively combining multimodal features and accurately embedding items. Our study specifically concentrates on artwork recommendation. In artwork recommendation, the textual data such as titles and descriptions notably influence users' preferences. Our research approach involves constructing multimodal embeddings of artworks by integrating both images and titles as a fundamental step.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AN10114171
• 書誌情報: 研究報告情報基礎とアクセス技術（IFAT）; 巻 2023-IFAT-152, 号 5, p. 1-3, 発行日 2023-09-14
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8884
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会