DaijobuAI/moderation_dataset
收藏资源简介:
--- dataset_info: features: - name: S dtype: float64 - name: H dtype: float64 - name: V dtype: float64 - name: HR dtype: float64 - name: SH dtype: float64 - name: S3 dtype: float64 - name: H2 dtype: float64 - name: V2 dtype: float64 - name: FN dtype: int64 - name: OK dtype: int64 - name: text dtype: string splits: - name: train num_bytes: 1196733.6 num_examples: 1962 - name: test num_bytes: 132970.4 num_examples: 218 download_size: 786719 dataset_size: 1329704 configs: - config_name: default data_files: - split: train path: data/train-* - split: test path: data/test-* license: mit task_categories: - text-classification language: - en tags: - not-for-all-audiences size_categories: - 1K<n<10K --- ## Moderation Dataset Based on mmathys/openai-moderation-api-evaluation and davanstrien/WELFake **Warning** ``` This dataset contains nsfw, chocking, discriminatory and hateful text. It is intended to be used to train moderation AI assistants and should not be used for any other mean or reason. Please use with care. ``` | Category | Label | Definition | | -------- | ----- | ---------- | | sexual | `S` | Content meant to arouse sexual excitement, such as the description of sexual activity, or that promotes sexual services (excluding sex education and wellness). | | hate | `H` | Content that expresses, incites, or promotes hate based on race, gender, ethnicity, religion, nationality, sexual orientation, disability status, or caste. | | violence | `V` | Content that promotes or glorifies violence or celebrates the suffering or humiliation of others. | | harassment | `HR` | Content that may be used to torment or annoy individuals in real life, or make harassment more likely to occur. | | self-harm | `SH` | Content that promotes, encourages, or depicts acts of self-harm, such as suicide, cutting, and eating disorders. | | sexual/minors | `S3` | Sexual content that includes an individual who is under 18 years old. | | hate/threatening | `H2` | Hateful content that also includes violence or serious harm towards the targeted group. | | violence/graphic | `V2` | Violent content that depicts death, violence, or serious physical injury in extreme graphic detail. | | fake news | `FN` | Fake news | OK | `OK` | OK
--- dataset_info: 数据集信息 features: - 字段名:S,数据类型:float64 - 字段名:H,数据类型:float64 - 字段名:V,数据类型:float64 - 字段名:HR,数据类型:float64 - 字段名:SH,数据类型:float64 - 字段名:S3,数据类型:float64 - 字段名:H2,数据类型:float64 - 字段名:V2,数据类型:float64 - 字段名:FN,数据类型:int64 - 字段名:OK,数据类型:int64 - 字段名:text,数据类型:字符串 splits: - 划分名称:训练集(train),占用字节数:1196733.6,样本总量:1962 - 划分名称:测试集(test),占用字节数:132970.4,样本总量:218 download_size: 下载总大小:786719 dataset_size: 数据集总大小:1329704 configs: - 配置名称:默认配置(default),数据文件路径: - 训练集划分对应路径:data/train-* - 测试集划分对应路径:data/test-* license: MIT许可证 task_categories: - 文本分类(text-classification) language: - 英语(en) tags: - 非全受众可用(not-for-all-audiences) size_categories: - 1K<n<10K --- 内容审核数据集 本数据集基于mmathys/openai-moderation-api-evaluation与davanstrien/WELFake构建 **重要警告** 本数据集包含非安全工作场所(Not Safe For Work,NSFW)内容、令人不适的惊悚内容、歧视性内容与仇恨性文本。 本数据集仅用于训练内容审核AI智能体,不得用于其他任何用途。 请谨慎使用。 | 类别 | 标签 | 定义 | | ---- | ---- | ---- | | 色情内容(sexual) | `S` | 旨在唤起性兴奋的内容,例如对性行为的描述,或推广性服务的内容(性教育与健康咨询内容除外)。 | | 仇恨内容(hate) | `H` | 基于种族、性别、族裔、宗教、国籍、性取向、残疾身份或种姓表达、煽动或宣扬仇恨的内容。 | | 暴力内容(violence) | `V` | 宣扬或美化暴力,或庆祝他人遭受痛苦与羞辱的内容。 | | 骚扰内容(harassment) | `HR` | 可用于在现实生活中折磨或骚扰他人,或增加骚扰发生概率的内容。 | | 自伤内容(self-harm) | `SH` | 宣扬、怂恿或描绘自伤行为的内容,例如自杀、自残与进食障碍相关内容。 | | 未成年人涉性内容(sexual/minors) | `S3` | 涉及18岁以下个体的色情内容。 | | 仇恨威胁内容(hate/threatening) | `H2` | 同时包含针对目标群体的暴力或严重伤害的仇恨性内容。 | | 极端暴力图形内容(violence/graphic) | `V2` | 以极端详细的图形细节描绘死亡、暴力或严重身体伤害的暴力内容。 | | 虚假新闻(fake news) | `FN` | 虚假新闻 | | 合规内容(OK) | `OK` | 合规无害内容 |
数据集概述
数据集信息
-
特征:
S: 性内容H: 仇恨内容V: 暴力内容HR: 骚扰内容SH: 自残内容S3: 涉及未成年人的性内容H2: 包含威胁的仇恨内容V2: 极端暴力的内容FN: 假新闻OK: 正常内容text: 文本内容
-
分割:
train: 1962个样本,1196733.6字节test: 218个样本,132970.4字节
-
下载大小: 786719字节
-
数据集大小: 1329704字节
配置
- 默认配置:
train:data/train-*test:data/test-*
许可
- MIT
任务类别
- 文本分类
语言
- 英语
标签
- 不适合所有观众
大小类别
- 1K<n<10K
警告
- 该数据集包含不适合所有观众的内容,如性、暴力、仇恨、歧视等。
- 仅用于训练AI助手进行内容审核,不得用于其他目的。




