HYUKI5/mteb__multi-hatecheck

mteb/multi-hatecheck（複製）

データの内容

11 言語のヘイトスピーチ検出の機能テストを、埋め込みモデルの評価タスク（MultiHateClassification）用にまとめたデータ。1 行が 1 件の合成文で、text に文、is_hateful に正解のラベル、lang に言語、functionality に検証する機能が入る。25 種類以上のヘイトの型と、判定の難しい非ヘイトの文を対象にしている。

データの概要

項目	値
行数	40,310（1 構成・1 分割の合計）
列数	4
容量	371.16 KB（ファイル 15 本）
ファイルの形式	jsonl.gz 11、yaml 1、py 1
言語	ara, cmn, deu, eng, fra, hin, ita, nld, pol, por, spa
公開	public

構成と分割

構成（config）	分割（split）ごとの行数	列数
default	test 40,310	4

列（構成 default・分割 test）

列	型	先頭の行の値
functionality	string	derog_neg_emote_h
text	string	انا بكره النساء.
is_hateful	string	hateful
lang	string	ara

行数と列は HF のデータセットビューアの集計（元のデータセットの現在の版で数えたもの）。全データセットの一覧は HYUKI5/dataset-catalog にある。

複製について

mteb/multi-hatecheck をリビジョン 5c7cf1c4a8fb7263c85e3df0a755f33878ae38c4 の時点でそのまま複製したもの。中身は変えていない。判断モデル Laya（convaiinnovations/laya）の微調整に使う候補として、調査メモ（Laya 学習用データセット調査）に載せたデータを集めている。

出典

• 元のデータセット: https://huggingface.co/datasets/mteb/multi-hatecheck
• 複製したリビジョン: 5c7cf1c4a8fb7263c85e3df0a755f33878ae38c4（2026-02-24 21:18:26+00:00）
• 元のライセンス表記: cc-by-4.0

論文

• https://arxiv.org/abs/2502.13595
• https://arxiv.org/abs/2210.07316
• https://aclanthology.org/2022.woah-1.15/
• https://aclanthology.org/2021.acl-long.4
• https://aclanthology.org/2022.woah-1.15

アノテーションの方法

（元のカードに該当の見出しが無いので、ラベルの付け方に触れた段落を抜き出した）

@inproceedings{rottger-etal-2021-hatecheck, abstract = {Detecting online hate is a difficult task that even state-of-the-art models struggle with. Typically, hate speech detection models are evaluated by measuring their performance on held-out test data using metrics such as accuracy and F1 score. However, this approach makes it difficult to identify specific model weak points. It also risks overestimating generalisable model performance due to increasingly well-evidenced systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, we introduce HateCheck, a suite of functional tests for hate speech detection models. We specify 29 model functionalities motivated by a review of previous research and a series of interviews with civil society stakeholders. We craft test cases for each functionality and validate their quality through a structured annotation process. To illustrate HateCheck{'}s utility, we test near-state-of-the-art transformer models as well as two popular commercial models, revealing critical model weaknesses.}, address = {Online}, author = {R{"o}ttger, Paul and Vidgen, Bertie and Nguyen, Dong and Waseem, Zeerak and Margetts, Helen and Pierrehumbert, Janet}, booktitle = {Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)}, doi = {10.18653/v1/2021.acl-long.4}, editor = {Zong, Chengqing and Xia, Fei and Li, Wenjie and Navigli, Roberto}, month = aug, pages = {41--58}, publisher = {Association for Computational Linguistics}, title = {{H}ate{C}heck: Functional Tests for Hate Speech Detection Models}, url = {https://aclanthology.org/2021.acl-long.4}, year = {2021}, }

---

このデータセットカードは MTEB を使って自動生成された。

微調整に使うときの注意

• 調査メモの台帳での区分: A（許諾型のライセンスだけで、疑義の注記が無い）。区分は記載の文言で機械的に付けたもので、法的な判定ではない。学習に入れる前に元のカードとライセンスの原文で確かめる。
• Laya の入力は 2,048 token まで（ModernBERT-large / mmBERT）。長い本文は先頭から黙って切られるので、切れる行は事前に除くか、要約・切り出しをしてから入れる。
• 分割は記事・文書単位で行い、同じ本文から作った別の問いを学習側と較正側に分けない。温度と閾値は学習に使っていない分割で別々に当てる。

調査メモの記述

メモ	節	ライセンス（記載）	確認	注記
multilingual_cjk.md	1. 大規模多言語データセット(XNLI, PAWS-X, MASSIVE, SIB-200, Belebele ほか) (L28)	cc-by-4.0	確認済	確認印なし(各ファイルの凡例により確認済扱い) ／ 原文注記: 機能テスト用の合成文で規模は小(10K 未満〜100K 未満)

元のカード（日本語訳）

以下は元のデータセットのカードの本文を日本語に訳したもの（機械翻訳）。訳に疑問があるときは、末尾の原文を確かめる。

---

MultiHateClassification

MTEB データセット

Massive Text Embedding Benchmark

ヘイトスピーチ検出のデータセットで、ヘイト（憎悪的）と非ヘイトの 2 値ラベルを持つ。25 種類以上の異なる種類のヘイトと、判定の難しい非ヘイトを含み、11 言語にわたる。

Task category	t2c
Domains	作成されたもの, 書き言葉
Reference	https://aclanthology.org/2022.woah-1.15/

このタスクでの評価方法

このタスクで埋め込みモデルを評価するには、次のコードを使う:

import mteb

task = mteb.get_tasks(["MultiHateClassification"])
evaluator = mteb.MTEB(task)

model = mteb.get_model(YOUR_MODEL)
evaluator.run(model)

mteb のタスクでモデルを実行する方法については、GitHub リポジトリ を参照のこと。

引用

このデータセットを使う場合は、データセットと mteb の両方を引用すること。このデータセットは MMTEB Contribution の一部として追加の処理を受けている可能性があるためである。

@inproceedings{rottger-etal-2021-hatecheck,
  abstract = {Detecting online hate is a difficult task that even state-of-the-art models struggle with. Typically, hate speech detection models are evaluated by measuring their performance on held-out test data using metrics such as accuracy and F1 score. However, this approach makes it difficult to identify specific model weak points. It also risks overestimating generalisable model performance due to increasingly well-evidenced systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, we introduce HateCheck, a suite of functional tests for hate speech detection models. We specify 29 model functionalities motivated by a review of previous research and a series of interviews with civil society stakeholders. We craft test cases for each functionality and validate their quality through a structured annotation process. To illustrate HateCheck{'}s utility, we test near-state-of-the-art transformer models as well as two popular commercial models, revealing critical model weaknesses.},
  address = {Online},
  author = {R{\"o}ttger, Paul  and
Vidgen, Bertie  and
Nguyen, Dong  and
Waseem, Zeerak  and
Margetts, Helen  and
Pierrehumbert, Janet},
  booktitle = {Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)},
  doi = {10.18653/v1/2021.acl-long.4},
  editor = {Zong, Chengqing  and
Xia, Fei  and
Li, Wenjie  and
Navigli, Roberto},
  month = aug,
  pages = {41--58},
  publisher = {Association for Computational Linguistics},
  title = {{H}ate{C}heck: Functional Tests for Hate Speech Detection Models},
  url = {https://aclanthology.org/2021.acl-long.4},
  year = {2021},
}

@inproceedings{rottger-etal-2022-multilingual,
  abstract = {Hate speech detection models are typically evaluated on held-out test sets. However, this risks painting an incomplete and potentially misleading picture of model performance because of increasingly well-documented systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, recent research has thus introduced functional tests for hate speech detection models. However, these tests currently only exist for English-language content, which means that they cannot support the development of more effective models in other languages spoken by billions across the world. To help address this issue, we introduce Multilingual HateCheck (MHC), a suite of functional tests for multilingual hate speech detection models. MHC covers 34 functionalities across ten languages, which is more languages than any other hate speech dataset. To illustrate MHC{'}s utility, we train and test a high-performing multilingual hate speech detection model, and reveal critical model weaknesses for monolingual and cross-lingual applications.},
  address = {Seattle, Washington (Hybrid)},
  author = {R{\"o}ttger, Paul  and
Seelawi, Haitham  and
Nozza, Debora  and
Talat, Zeerak  and
Vidgen, Bertie},
  booktitle = {Proceedings of the Sixth Workshop on Online Abuse and Harms (WOAH)},
  doi = {10.18653/v1/2022.woah-1.15},
  editor = {Narang, Kanika  and
Mostafazadeh Davani, Aida  and
Mathias, Lambert  and
Vidgen, Bertie  and
Talat, Zeerak},
  month = jul,
  pages = {154--169},
  publisher = {Association for Computational Linguistics},
  title = {Multilingual {H}ate{C}heck: Functional Tests for Multilingual Hate Speech Detection Models},
  url = {https://aclanthology.org/2022.woah-1.15},
  year = {2022},
}


@article{enevoldsen2025mmtebmassivemultilingualtext,
  title={MMTEB: Massive Multilingual Text Embedding Benchmark},
  author={Kenneth Enevoldsen and Isaac Chung and Imene Kerboua and Márton Kardos and Ashwin Mathur and David Stap and Jay Gala and Wissam Siblini and Dominik Krzemiński and Genta Indra Winata and Saba Sturua and Saiteja Utpala and Mathieu Ciancone and Marion Schaeffer and Gabriel Sequeira and Diganta Misra and Shreeya Dhakal and Jonathan Rystrøm and Roman Solomatin and Ömer Çağatan and Akash Kundu and Martin Bernstorff and Shitao Xiao and Akshita Sukhlecha and Bhavish Pahwa and Rafał Poświata and Kranthi Kiran GV and Shawon Ashraf and Daniel Auras and Björn Plüster and Jan Philipp Harries and Loïc Magne and Isabelle Mohr and Mariya Hendriksen and Dawei Zhu and Hippolyte Gisserot-Boukhlef and Tom Aarsen and Jan Kostkan and Konrad Wojtasik and Taemin Lee and Marek Šuppa and Crystina Zhang and Roberta Rocca and Mohammed Hamdy and Andrianos Michail and John Yang and Manuel Faysse and Aleksei Vatolin and Nandan Thakur and Manan Dey and Dipam Vasani and Pranjal Chitale and Simone Tedeschi and Nguyen Tai and Artem Snegirev and Michael Günther and Mengzhou Xia and Weijia Shi and Xing Han Lù and Jordan Clive and Gayatri Krishnakumar and Anna Maksimova and Silvan Wehrli and Maria Tikhonova and Henil Panchal and Aleksandr Abramov and Malte Ostendorff and Zheng Liu and Simon Clematide and Lester James Miranda and Alena Fenogenova and Guangyu Song and Ruqiya Bin Safi and Wen-Ding Li and Alessia Borghini and Federico Cassano and Hongjin Su and Jimmy Lin and Howard Yen and Lasse Hansen and Sara Hooker and Chenghao Xiao and Vaibhav Adlakha and Orion Weller and Siva Reddy and Niklas Muennighoff},
  publisher = {arXiv},
  journal={arXiv preprint arXiv:2502.13595},
  year={2025},
  url={https://arxiv.org/abs/2502.13595},
  doi = {10.48550/arXiv.2502.13595},
}

@article{muennighoff2022mteb,
  author = {Muennighoff, Niklas and Tazi, Nouamane and Magne, Lo{\"\i}c and Reimers, Nils},
  title = {MTEB: Massive Text Embedding Benchmark},
  publisher = {arXiv},
  journal={arXiv preprint arXiv:2210.07316},
  year = {2022}
  url = {https://arxiv.org/abs/2210.07316},
  doi = {10.48550/ARXIV.2210.07316},
}

データセットの統計

データセットの統計

次のコードには、タスクの記述統計が含まれている。これらは次の方法でも取得できる:

import mteb

task = mteb.get_task("MultiHateClassification")

desc_stats = task.metadata.descriptive_stats

{
    "test": {
        "num_samples": 11000,
        "number_of_characters": 502013,
        "number_texts_intersect_with_train": 16,
        "min_text_length": 1,
        "average_text_length": 45.63754545454545,
        "max_text_length": 135,
        "unique_text": 10990,
        "unique_labels": 2,
        "labels": {
            "0": {
                "count": 7661
            },
            "1": {
                "count": 3339
            }
        }
    },
    "train": {
        "num_samples": 11000,
        "number_of_characters": 505993,
        "number_texts_intersect_with_train": null,
        "min_text_length": 4,
        "average_text_length": 45.99936363636364,
        "max_text_length": 131,
        "unique_text": 10993,
        "unique_labels": 2,
        "labels": {
            "0": {
                "count": 7659
            },
            "1": {
                "count": 3341
            }
        }
    }
}

---

このデータセットカードは MTEB を使って自動生成された。

原文（英語）

MultiHateClassification

An MTEB dataset

Massive Text Embedding Benchmark

Hate speech detection dataset with binary (hateful vs non-hateful) labels. Includes 25+ distinct types of hate and challenging non-hate, and 11 languages.

Task category	t2c
Domains	Constructed, Written
Reference	https://aclanthology.org/2022.woah-1.15/

How to evaluate on this task

You can evaluate an embedding model on this dataset using the following code:

import mteb

task = mteb.get_tasks(["MultiHateClassification"])
evaluator = mteb.MTEB(task)

model = mteb.get_model(YOUR_MODEL)
evaluator.run(model)

To learn more about how to run models on mteb task check out the GitHub repitory.

Citation

If you use this dataset, please cite the dataset as well as mteb, as this dataset likely includes additional processing as a part of the MMTEB Contribution.

@inproceedings{rottger-etal-2021-hatecheck,
  abstract = {Detecting online hate is a difficult task that even state-of-the-art models struggle with. Typically, hate speech detection models are evaluated by measuring their performance on held-out test data using metrics such as accuracy and F1 score. However, this approach makes it difficult to identify specific model weak points. It also risks overestimating generalisable model performance due to increasingly well-evidenced systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, we introduce HateCheck, a suite of functional tests for hate speech detection models. We specify 29 model functionalities motivated by a review of previous research and a series of interviews with civil society stakeholders. We craft test cases for each functionality and validate their quality through a structured annotation process. To illustrate HateCheck{'}s utility, we test near-state-of-the-art transformer models as well as two popular commercial models, revealing critical model weaknesses.},
  address = {Online},
  author = {R{\"o}ttger, Paul  and
Vidgen, Bertie  and
Nguyen, Dong  and
Waseem, Zeerak  and
Margetts, Helen  and
Pierrehumbert, Janet},
  booktitle = {Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)},
  doi = {10.18653/v1/2021.acl-long.4},
  editor = {Zong, Chengqing  and
Xia, Fei  and
Li, Wenjie  and
Navigli, Roberto},
  month = aug,
  pages = {41--58},
  publisher = {Association for Computational Linguistics},
  title = {{H}ate{C}heck: Functional Tests for Hate Speech Detection Models},
  url = {https://aclanthology.org/2021.acl-long.4},
  year = {2021},
}

@inproceedings{rottger-etal-2022-multilingual,
  abstract = {Hate speech detection models are typically evaluated on held-out test sets. However, this risks painting an incomplete and potentially misleading picture of model performance because of increasingly well-documented systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, recent research has thus introduced functional tests for hate speech detection models. However, these tests currently only exist for English-language content, which means that they cannot support the development of more effective models in other languages spoken by billions across the world. To help address this issue, we introduce Multilingual HateCheck (MHC), a suite of functional tests for multilingual hate speech detection models. MHC covers 34 functionalities across ten languages, which is more languages than any other hate speech dataset. To illustrate MHC{'}s utility, we train and test a high-performing multilingual hate speech detection model, and reveal critical model weaknesses for monolingual and cross-lingual applications.},
  address = {Seattle, Washington (Hybrid)},
  author = {R{\"o}ttger, Paul  and
Seelawi, Haitham  and
Nozza, Debora  and
Talat, Zeerak  and
Vidgen, Bertie},
  booktitle = {Proceedings of the Sixth Workshop on Online Abuse and Harms (WOAH)},
  doi = {10.18653/v1/2022.woah-1.15},
  editor = {Narang, Kanika  and
Mostafazadeh Davani, Aida  and
Mathias, Lambert  and
Vidgen, Bertie  and
Talat, Zeerak},
  month = jul,
  pages = {154--169},
  publisher = {Association for Computational Linguistics},
  title = {Multilingual {H}ate{C}heck: Functional Tests for Multilingual Hate Speech Detection Models},
  url = {https://aclanthology.org/2022.woah-1.15},
  year = {2022},
}


@article{enevoldsen2025mmtebmassivemultilingualtext,
  title={MMTEB: Massive Multilingual Text Embedding Benchmark},
  author={Kenneth Enevoldsen and Isaac Chung and Imene Kerboua and Márton Kardos and Ashwin Mathur and David Stap and Jay Gala and Wissam Siblini and Dominik Krzemiński and Genta Indra Winata and Saba Sturua and Saiteja Utpala and Mathieu Ciancone and Marion Schaeffer and Gabriel Sequeira and Diganta Misra and Shreeya Dhakal and Jonathan Rystrøm and Roman Solomatin and Ömer Çağatan and Akash Kundu and Martin Bernstorff and Shitao Xiao and Akshita Sukhlecha and Bhavish Pahwa and Rafał Poświata and Kranthi Kiran GV and Shawon Ashraf and Daniel Auras and Björn Plüster and Jan Philipp Harries and Loïc Magne and Isabelle Mohr and Mariya Hendriksen and Dawei Zhu and Hippolyte Gisserot-Boukhlef and Tom Aarsen and Jan Kostkan and Konrad Wojtasik and Taemin Lee and Marek Šuppa and Crystina Zhang and Roberta Rocca and Mohammed Hamdy and Andrianos Michail and John Yang and Manuel Faysse and Aleksei Vatolin and Nandan Thakur and Manan Dey and Dipam Vasani and Pranjal Chitale and Simone Tedeschi and Nguyen Tai and Artem Snegirev and Michael Günther and Mengzhou Xia and Weijia Shi and Xing Han Lù and Jordan Clive and Gayatri Krishnakumar and Anna Maksimova and Silvan Wehrli and Maria Tikhonova and Henil Panchal and Aleksandr Abramov and Malte Ostendorff and Zheng Liu and Simon Clematide and Lester James Miranda and Alena Fenogenova and Guangyu Song and Ruqiya Bin Safi and Wen-Ding Li and Alessia Borghini and Federico Cassano and Hongjin Su and Jimmy Lin and Howard Yen and Lasse Hansen and Sara Hooker and Chenghao Xiao and Vaibhav Adlakha and Orion Weller and Siva Reddy and Niklas Muennighoff},
  publisher = {arXiv},
  journal={arXiv preprint arXiv:2502.13595},
  year={2025},
  url={https://arxiv.org/abs/2502.13595},
  doi = {10.48550/arXiv.2502.13595},
}

@article{muennighoff2022mteb,
  author = {Muennighoff, Niklas and Tazi, Nouamane and Magne, Lo{\"\i}c and Reimers, Nils},
  title = {MTEB: Massive Text Embedding Benchmark},
  publisher = {arXiv},
  journal={arXiv preprint arXiv:2210.07316},
  year = {2022}
  url = {https://arxiv.org/abs/2210.07316},
  doi = {10.48550/ARXIV.2210.07316},
}

Dataset Statistics

Dataset Statistics

The following code contains the descriptive statistics from the task. These can also be obtained using:

import mteb

task = mteb.get_task("MultiHateClassification")

desc_stats = task.metadata.descriptive_stats

{
    "test": {
        "num_samples": 11000,
        "number_of_characters": 502013,
        "number_texts_intersect_with_train": 16,
        "min_text_length": 1,
        "average_text_length": 45.63754545454545,
        "max_text_length": 135,
        "unique_text": 10990,
        "unique_labels": 2,
        "labels": {
            "0": {
                "count": 7661
            },
            "1": {
                "count": 3339
            }
        }
    },
    "train": {
        "num_samples": 11000,
        "number_of_characters": 505993,
        "number_texts_intersect_with_train": null,
        "min_text_length": 4,
        "average_text_length": 45.99936363636364,
        "max_text_length": 131,
        "unique_text": 10993,
        "unique_labels": 2,
        "labels": {
            "0": {
                "count": 7659
            },
            "1": {
                "count": 3341
            }
        }
    }
}

---

This dataset card was automatically generated using MTEB