@inproceedings{feng-etal-2026-constructive,
title = "Constructive Comment Extraction from {J}apanese {Y}ou{T}ube Shorts: A Dataset and Empirical Study",
author = "Feng, Siyuan and
Wang, Luecheng and
Ogura, Koyo and
Iwamoto, Yuma and
Sato, Seinosuke and
Kometani, Yuki and
Itabashi, Masaki and
Ozaki, Shintaro and
Sakai, Yusuke and
Matsuno, Shogo and
Kamigaito, Hidetaka and
Hayashi, Katsuhiko and
Obara, Ryoma and
Takase, Sho and
Watanabe, Taro",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the {A}sia-Pacific Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 2: Short Papers)",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.aacl-short.3/",
pages = "22--34",
ISBN = "979-8-89176-440-8",
abstract = "Identifying constructive comments from user-generated content is important for improving information discovery and discussion quality on platforms such as YouTube. However, existing resources focus predominantly on English news comments, leaving short-form video platforms underexplored despite their distinct characteristics: comments are shorter, more reactive, and not well captured by popularity signals. In this work, we introduce the first Japanese YouTube Shorts dataset for constructive comment extraction, featuring multi-annotator labels that preserve naturally occurring disagreement. We also develop a real-time system that surfaces constructive comments from a noisy comment stream, evaluated in a preliminary user study. To train under noisy annotation conditions, we combine self-training with high-confidence pseudo-labels and active learning. Experiments show that iterative data expansion improves performance, while our compact task-specific model achieves the highest PR-AUC point estimate among the evaluated systems. These findings demonstrate the value of dataset construction and targeted training for constructive comment extraction in realistic settings. Our dataset and code are available at https://github.com/nlp-hayashi/japanese-youtube-constructive-comments"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="feng-etal-2026-constructive">
<titleInfo>
<title>Constructive Comment Extraction from Japanese YouTube Shorts: A Dataset and Empirical Study</title>
</titleInfo>
<name type="personal">
<namePart type="given">Siyuan</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Luecheng</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Koyo</namePart>
<namePart type="family">Ogura</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuma</namePart>
<namePart type="family">Iwamoto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seinosuke</namePart>
<namePart type="family">Sato</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuki</namePart>
<namePart type="family">Kometani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Masaki</namePart>
<namePart type="family">Itabashi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shintaro</namePart>
<namePart type="family">Ozaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yusuke</namePart>
<namePart type="family">Sakai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shogo</namePart>
<namePart type="family">Matsuno</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hidetaka</namePart>
<namePart type="family">Kamigaito</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Katsuhiko</namePart>
<namePart type="family">Hayashi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ryoma</namePart>
<namePart type="family">Obara</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sho</namePart>
<namePart type="family">Takase</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Taro</namePart>
<namePart type="family">Watanabe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 2: Short Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-440-8</identifier>
</relatedItem>
<abstract>Identifying constructive comments from user-generated content is important for improving information discovery and discussion quality on platforms such as YouTube. However, existing resources focus predominantly on English news comments, leaving short-form video platforms underexplored despite their distinct characteristics: comments are shorter, more reactive, and not well captured by popularity signals. In this work, we introduce the first Japanese YouTube Shorts dataset for constructive comment extraction, featuring multi-annotator labels that preserve naturally occurring disagreement. We also develop a real-time system that surfaces constructive comments from a noisy comment stream, evaluated in a preliminary user study. To train under noisy annotation conditions, we combine self-training with high-confidence pseudo-labels and active learning. Experiments show that iterative data expansion improves performance, while our compact task-specific model achieves the highest PR-AUC point estimate among the evaluated systems. These findings demonstrate the value of dataset construction and targeted training for constructive comment extraction in realistic settings. Our dataset and code are available at https://github.com/nlp-hayashi/japanese-youtube-constructive-comments</abstract>
<identifier type="citekey">feng-etal-2026-constructive</identifier>
<location>
<url>https://aclanthology.org/2026.aacl-short.3/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>22</start>
<end>34</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Constructive Comment Extraction from Japanese YouTube Shorts: A Dataset and Empirical Study
%A Feng, Siyuan
%A Wang, Luecheng
%A Ogura, Koyo
%A Iwamoto, Yuma
%A Sato, Seinosuke
%A Kometani, Yuki
%A Itabashi, Masaki
%A Ozaki, Shintaro
%A Sakai, Yusuke
%A Matsuno, Shogo
%A Kamigaito, Hidetaka
%A Hayashi, Katsuhiko
%A Obara, Ryoma
%A Takase, Sho
%A Watanabe, Taro
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 2: Short Papers)
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-440-8
%F feng-etal-2026-constructive
%X Identifying constructive comments from user-generated content is important for improving information discovery and discussion quality on platforms such as YouTube. However, existing resources focus predominantly on English news comments, leaving short-form video platforms underexplored despite their distinct characteristics: comments are shorter, more reactive, and not well captured by popularity signals. In this work, we introduce the first Japanese YouTube Shorts dataset for constructive comment extraction, featuring multi-annotator labels that preserve naturally occurring disagreement. We also develop a real-time system that surfaces constructive comments from a noisy comment stream, evaluated in a preliminary user study. To train under noisy annotation conditions, we combine self-training with high-confidence pseudo-labels and active learning. Experiments show that iterative data expansion improves performance, while our compact task-specific model achieves the highest PR-AUC point estimate among the evaluated systems. These findings demonstrate the value of dataset construction and targeted training for constructive comment extraction in realistic settings. Our dataset and code are available at https://github.com/nlp-hayashi/japanese-youtube-constructive-comments
%U https://aclanthology.org/2026.aacl-short.3/
%P 22-34
Markdown (Informal)
[Constructive Comment Extraction from Japanese YouTube Shorts: A Dataset and Empirical Study](https://aclanthology.org/2026.aacl-short.3/) (Feng et al., AACL-IJCNLP 2026)
ACL
- Siyuan Feng, Luecheng Wang, Koyo Ogura, Yuma Iwamoto, Seinosuke Sato, Yuki Kometani, Masaki Itabashi, Shintaro Ozaki, Yusuke Sakai, Shogo Matsuno, Hidetaka Kamigaito, Katsuhiko Hayashi, Ryoma Obara, Sho Takase, and Taro Watanabe. 2026. Constructive Comment Extraction from Japanese YouTube Shorts: A Dataset and Empirical Study. In Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 2: Short Papers), pages 22–34, Hengqin, China. Association for Computational Linguistics.