PolyAI/minds14
收藏资源简介：
MInDS-14 是一个用于口语数据意图检测任务训练和评估的资源。它涵盖了从电子银行领域的商业系统中提取的 14 个意图，以及 14 种不同语言变体的口语示例。该数据集支持多种语言，包括英语、法语、意大利语、西班牙语、葡萄牙语、德语、荷兰语、俄语、波兰语、捷克语、韩语和中文。数据集的许可证为 CC BY 4.0。数据集的大小在 10K 到 100K 之间。主要任务是自动语音识别和关键词检测。数据集的配置包括 all 和各个语言的具体配置。每个配置都包括训练集的路径、音频文件、转录文本、意图类别和语言 ID。数据集还提供了详细的文件大小和示例数量信息。
MInDS-14 is a training and evaluation resource for intent detection tasks with spoken data. It covers 14 intents extracted from a commercial system in the e-banking domain, associated with spoken examples in 14 diverse language varieties. The dataset supports multiple languages, including English, French, Italian, Spanish, Portuguese, German, Dutch, Russian, Polish, Czech, Korean, and Chinese. The license for the dataset is CC BY 4.0. The dataset size falls within the category of 10K < n < 100K. The main tasks include automatic speech recognition and keyword spotting. The dataset configurations include all and specific configurations for each language. Each configuration includes paths, audio files, transcriptions, intent classes, and language IDs for the training set. The dataset also provides detailed information about file sizes and the number of examples.
MInDS-14 数据集概述
数据集描述
- 数据集名称: MInDS-14
- 数据集用途: 用于意图检测任务的语音数据训练和评估
- 覆盖意图数量: 14个
- 语言种类: 14种多样化的语言变体
- 总磁盘使用量: 约500 MB
数据集结构
数据实例
- 配置:
fr-FR- 下载的数据集文件大小: 471 MB
- 生成的数据集大小: 300 KB
- 总磁盘使用量: 471 MB
示例数据实例
json { "path": "/home/patrick/.cache/huggingface/datasets/downloads/extracted/3ebe2265b2f102203be5e64fa8e533e0c6742e72268772c8ac1834c5a1a921e3/fr-FR~ADDRESS/response_4.wav", "audio": { "path": "/home/patrick/.cache/huggingface/datasets/downloads/extracted/3ebe2265b2f102203be5e64fa8e533e0c6742e72268772c8ac1834c5a1a921e3/fr-FR~ADDRESS/response_4.wav", "array": array( [0.0, 0.0, 0.0, ..., 0.0, 0.00048828, -0.00024414], dtype=float32 ), "sampling_rate": 8000, }, "transcription": "je souhaite changer mon adresse", "english_transcription": "I want to change my address", "intent_class": 1, "lang_id": 6, }
数据字段
- path (str): 音频文件路径
- audio (dict): 包含音频数组、采样率和音频路径的音频对象
- transcription (str): 音频文件的转录文本
- english_transcription (str): 音频文件的英文转录文本
- intent_class (int): 意图类别的ID
- lang_id (int): 语言的ID
数据分割
- 每个配置仅包含
"train"分割，包含约600个示例。
数据集创建
- 标注创建者:
- 专家生成
- 众包
- 机器生成
- 语言创建者:
- 众包
- 专家生成
- 语言:
- en, fr, it, es, pt, de, nl, ru, pl, cs, ko, zh
- 语言BCP47:
- en, en-GB, en-US, en-AU, fr, it, es, pt, de, nl, ru, pl, cs, ko, zh
- 许可证: CC-BY-4.0
- 多语言性: 多语言
- 数据集大小分类: 10K<n<100K
- 任务类别:
- 自动语音识别
- 语音处理
- 任务ID:
- 语音识别
- 关键词识别
额外信息
许可证信息
- 所有数据集均在 Creative Commons license (CC-BY) 下授权。
引用信息
bibtex @article{DBLP:journals/corr/abs-2104-08524, author = {Daniela Gerz and Pei{-}Hao Su and Razvan Kusztos and Avishek Mondal and Michal Lis and Eshan Singhal and Nikola Mrksic and Tsung{-}Hsien Wen and Ivan Vulic}, title = {Multilingual and Cross-Lingual Intent Detection from Spoken Data}, journal = {CoRR}, volume = {abs/2104.08524}, year = {2021}, url = {https://arxiv.org/abs/2104.08524}, eprinttype = {arXiv}, eprint = {2104.08524}, timestamp = {Mon, 26 Apr 2021 17:25:10 +0200}, biburl = {https://dblp.org/rec/journals/corr/abs-2104-08524.bib}, bibsource = {dblp computer science bibliography, https://dblp.org} }
