taskmaster2

# Dataset Card for Taskmaster-2

## Table of Contents
- [Dataset Description](#dataset-description)
- [Dataset Summary](#dataset-summary)
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
- [Languages](#languages)
- [Dataset Structure](#dataset-structure)
- [Data Instances](#data-instances)
- [Data Fields](#data-fields)
- [Data Splits](#data-splits)
- [Dataset Creation](#dataset-creation)
- [Curation Rationale](#curation-rationale)
- [Source Data](#source-data)
- [Annotations](#annotations)
- [Personal and Sensitive Information](#personal-and-sensitive-information)
- [Considerations for Using the Data](#considerations-for-using-the-data)
- [Social Impact of Dataset](#social-impact-of-dataset)
- [Discussion of Biases](#discussion-of-biases)
- [Other Known Limitations](#other-known-limitations)
- [Additional Information](#additional-information)
- [Dataset Curators](#dataset-curators)
- [Licensing Information](#licensing-information)
- [Citation Information](#citation-information)
- [Contributions](#contributions)

## Dataset Description

- **Homepage:** [Taskmaster-1](https://research.google/tools/datasets/taskmaster-1/)
- **Repository:** [GitHub](https://github.com/google-research-datasets/Taskmaster/tree/master/TM-2-2020)
- **Paper:** [Taskmaster-1: Toward a Realistic and Diverse Dialog Dataset](https://arxiv.org/abs/1909.05358)
- **Leaderboard:** N/A
- **Point of Contact:** [Taskmaster Googlegroup]([e-mail])

### Dataset Summary

Taskmaster is dataset for goal oriented conversations. The Taskmaster-2 dataset consists of 17,289 dialogs
in the seven domains which include restaurants, food ordering, movies, hotels, flights, music and sports.
Unlike Taskmaster-1, which includes both written "self-dialogs" and spoken two-person dialogs,
Taskmaster-2 consists entirely of spoken two-person dialogs. In addition, while Taskmaster-1 is
almost exclusively task-based, Taskmaster-2 contains a good number of search- and recommendation-oriented dialogs.
All dialogs in this release were created using a Wizard of Oz (WOz) methodology in which crowdsourced
workers played the role of a 'user' and trained call center operators played the role of the 'assistant'.
In this way, users were led to believe they were interacting with an automated system that “spoke”
using text-to-speech (TTS) even though it was in fact a human behind the scenes.
As a result, users could express themselves however they chose in the context of an automated interface.

### Supported Tasks and Leaderboards

[More Information Needed]

### Languages

The dataset is in English language.

## Dataset Structure

### Data Instances

A typical example looks like this

```
{
"conversation_id": "dlg-0047a087-6a3c-4f27-b0e6-268f53a2e013",
"instruction_id": "flight-6",
"utterances": [
{
"index": 0,
"segments": [],
"speaker": "USER",
"text": "Hi, I'm looking for a flight. I need to visit a friend."
},
{
"index": 1,
"segments": [],
"speaker": "ASSISTANT",
"text": "Hello, how can I help you?"
},
{
"index": 2,
"segments": [],
"speaker": "ASSISTANT",
"text": "Sure, I can help you with that."
},
{
"index": 3,
"segments": [],
"speaker": "ASSISTANT",
"text": "On what dates?"
},
{
"index": 4,
"segments": [
{
"annotations": [
{
"name": "flight_search.date.depart_origin"
}
],
"end_index": 37,
"start_index": 27,
"text": "March 20th"
},
{
"annotations": [
{
"name": "flight_search.date.return"
}
],
"end_index": 45,
"start_index": 41,
"text": "22nd"
}
],
"speaker": "USER",
"text": "I'm looking to travel from March 20th to 22nd."
}
]
}
```

### Data Fields

Each conversation in the data file has the following structure:

- `conversation_id`: A universally unique identifier with the prefix 'dlg-'. The ID has no meaning.
- `utterances`: A list of utterances that make up the conversation.
- `instruction_id`: A reference to the file(s) containing the user (and, if applicable, agent) instructions for this conversation.

Each utterance has the following fields:

- `index`: A 0-based index indicating the order of the utterances in the conversation.
- `speaker`: Either USER or ASSISTANT, indicating which role generated this utterance.
- `text`: The raw text of the utterance. In case of self dialogs (one_person_dialogs), this is written by the crowdsourced worker. In case of the WOz dialogs, 'ASSISTANT' turns are written and 'USER' turns are transcribed from the spoken recordings of crowdsourced workers.
- `segments`: A list of various text spans with semantic annotations.

Each segment has the following fields:

- `start_index`: The position of the start of the annotation in the utterance text.
- `end_index`: The position of the end of the annotation in the utterance text.
- `text`: The raw text that has been annotated.
- `annotations`: A list of annotation details for this segment.

Each annotation has a single field:

- `name`: The annotation name.

### Data Splits

There are no deafults splits for all the config. The below table lists the number of examples in each config.

| Config | Train |
|-------------------|--------|
| flights | 2481 |
| food-orderings | 1050 |
| hotels | 2355 |
| movies | 3047 |
| music | 1602 |
| restaurant-search | 3276 |
| sports | 3478 |

## Dataset Creation

### Curation Rationale

[More Information Needed]

### Source Data

[More Information Needed]

#### Initial Data Collection and Normalization

[More Information Needed]

#### Who are the source language producers?

[More Information Needed]

### Annotations

[More Information Needed]

#### Annotation process

[More Information Needed]

#### Who are the annotators?

[More Information Needed]

### Personal and Sensitive Information

[More Information Needed]

## Considerations for Using the Data

### Social Impact of Dataset

[More Information Needed]

### Discussion of Biases

[More Information Needed]

### Other Known Limitations

[More Information Needed]

## Additional Information

### Dataset Curators

[More Information Needed]

### Licensing Information

The dataset is licensed under `Creative Commons Attribution 4.0 License`

### Citation Information

[More Information Needed]
```
@inproceedings{48484,
title = {Taskmaster-1: Toward a Realistic and Diverse Dialog Dataset},
author = {Bill Byrne and Karthik Krishnamoorthi and Chinnadhurai Sankar and Arvind Neelakantan and Daniel Duckworth and Semih Yavuz and Ben Goodrich and Amit Dubey and Kyu-Young Kim and Andy Cedilnik},
year = {2019}
}
```
### Contributions

Thanks to [@patil-suraj](https://github.com/patil-suraj) for adding this dataset.

# 数据集卡片：Taskmaster-2

## 目录
- [数据集描述](#dataset-description)
- [数据集概述](#dataset-summary)
- [支持任务与基准测试榜](#supported-tasks-and-leaderboards)
- [语言](#languages)
- [数据集结构](#dataset-structure)
- [数据实例](#data-instances)
- [数据字段](#data-fields)
- [数据划分](#data-splits)
- [数据集构建](#dataset-creation)
- [构建初衷](#curation-rationale)
- [源数据](#source-data)
- [标注](#annotations)
- [个人与敏感信息](#personal-and-sensitive-information)
- [数据集使用注意事项](#considerations-for-using-the-data)
- [数据集的社会影响](#social-impact-of-dataset)
- [偏见分析](#discussion-of-biases)
- [其他已知局限](#other-known-limitations)
- [附加信息](#additional-information)
- [数据集维护者](#dataset-curators)
- [许可信息](#licensing-information)
- [引用信息](#citation-information)
- [贡献致谢](#contributions)

## 数据集描述

- **主页**：[Taskmaster-1](https://research.google/tools/datasets/taskmaster-1/)
- **代码仓库**：[GitHub](https://github.com/google-research-datasets/Taskmaster/tree/master/TM-2-2020)
- **相关论文**：[Taskmaster-1: Toward a Realistic and Diverse Dialog Dataset](https://arxiv.org/abs/1909.05358)
- **基准测试榜**：无
- **联系人**：[Taskmaster 工作组]([e-mail])

### 数据集概述

Taskmaster是面向目标的对话（goal-oriented conversations）数据集。Taskmaster-2数据集包含17289条对话，涵盖7个领域，分别为餐饮、食品订购、影视、酒店、航班、音乐与体育。与同时包含书面“自对话”与双人口语对话的Taskmaster-1不同，Taskmaster-2仅包含双人口语对话。此外，Taskmaster-1几乎完全基于任务型场景，而Taskmaster-2还包含大量搜索与推荐导向的对话。本版本发布的所有对话均采用绿野仙踪（Wizard of Oz, WOz）方法构建：众包工人扮演“用户”角色，经过培训的呼叫中心操作员扮演“助手”角色。在此设定下，用户会误以为自己在与采用文本转语音（text-to-speech, TTS）技术的自动化系统交互，而实际上后台由人类充当助手。因此，用户可在自动化交互界面中自由表达自身诉求。

### 支持任务与基准测试榜

[需补充更多信息]

### 语言

本数据集采用英语编写。

## 数据集结构

### 数据实例

典型的数据实例如下所示：

{
"conversation_id": "dlg-0047a087-6a3c-4f27-b0e6-268f53a2e013",
"instruction_id": "flight-6",
"utterances": [
{
"index": 0,
"segments": [],
"speaker": "USER",
"text": "Hi, I'm looking for a flight. I need to visit a friend."
},
{
"index": 1,
"segments": [],
"speaker": "ASSISTANT",
"text": "Hello, how can I help you?"
},
{
"index": 2,
"segments": [],
"speaker": "ASSISTANT",
"text": "Sure, I can help you with that."
},
{
"index": 3,
"segments": [],
"speaker": "ASSISTANT",
"text": "On what dates?"
},
{
"index": 4,
"segments": [
{
"annotations": [
{
"name": "flight_search.date.depart_origin"
}
],
"end_index": 37,
"start_index": 27,
"text": "March 20th"
},
{
"annotations": [
{
"name": "flight_search.date.return"
}
],
"end_index": 45,
"start_index": 41,
"text": "22nd"
}
],
"speaker": "USER",
"text": "I'm looking to travel from March 20th to 22nd."
}
]
}

### 数据字段

数据文件中的每条对话均遵循以下结构：
- `conversation_id`：以`dlg-`为前缀的全局唯一标识符，无实际语义。
- `utterances`：组成对话的话语列表。
- `instruction_id`：指向包含本次对话用户（及若适用的助手）指令的文件的引用。

每条话语包含以下字段：
- `index`：基于0的索引，用于标识话语在对话中的顺序。
- `speaker`：取值为`USER`或`ASSISTANT`，分别代表生成该话语的角色。
- `text`：话语的原始文本。若为自对话（one_person_dialogs），则由众包工人撰写；若为WOz对话，则助手（ASSISTANT）话语由人工撰写，用户（USER）话语由众包工人的口语录音转录而来。
- `segments`：带有语义标注的各类文本片段列表。

每个文本片段包含以下字段：
- `start_index`：标注在话语文本中的起始位置。
- `end_index`：标注在话语文本中的结束位置。
- `text`：被标注的原始文本。
- `annotations`：该片段的各类标注详情列表。

每个标注仅包含一个字段：
- `name`：标注名称。

### 数据划分

并非所有配置均存在默认划分。下表列出了各配置下的样本数量：

| 配置名称 | 训练集样本数 |
|---------------------|--------------|
| 航班查询 | 2481 |
| 食品订购 | 1050 |
| 酒店预订 | 2355 |
| 影视查询 | 3047 |
| 音乐查询 | 1602 |
| 餐饮搜索 | 3276 |
| 体育查询 | 3478 |

## 数据集构建

### 构建初衷

[需补充更多信息]

### 源数据

[需补充更多信息]

#### 初始数据收集与标准化

[需补充更多信息]

#### 源语言生成者是谁？

[需补充更多信息]

### 标注

[需补充更多信息]

#### 标注流程

[需补充更多信息]

#### 标注人员是谁？

[需补充更多信息]

### 个人与敏感信息

[需补充更多信息]

## 数据集使用注意事项

### 数据集的社会影响

[需补充更多信息]

### 偏见分析

[需补充更多信息]

### 其他已知局限

[需补充更多信息]

## 附加信息

### 数据集维护者

[需补充更多信息]

### 许可信息

本数据集采用`知识共享署名4.0许可（Creative Commons Attribution 4.0 License）`进行授权。

### 引用信息

[需补充更多信息]

@inproceedings{48484,
title = {Taskmaster-1: Toward a Realistic and Diverse Dialog Dataset},
author = {Bill Byrne and Karthik Krishnamoorthi and Chinnadhurai Sankar and Arvind Neelakantan and Daniel Duckworth and Semih Yavuz and Ben Goodrich and Amit Dubey and Kyu-Young Kim and Andy Cedilnik},
year = {2019}
}

### 贡献致谢

感谢 [@patil-suraj](https://github.com/patil-suraj) 为本数据集的收录提供支持。