数据集说明 | Dataset Description
本数据集是古代汉语历史事件检测数据集（CHED），旨在帮助研究者对古文历史文本中的事件进行检测与分类。数据集涵盖了《二十四史》中多个历史人物和事件的记录，适用于构建和评估古代历史文本事件检测的算法与模型。
This dataset, named the Classical Chinese Historical Event Dataset (CHED), is designed to aid researchers in detecting and classifying events within classical Chinese historical texts. It covers multiple historical figures and events recorded in the "Twenty-Four Histories," making it suitable for constructing and evaluating algorithms and models for event detection in ancient historical texts.
数据集文件 | Dataset Files
| 文件名 | 描述 |
|---|---|
| train.jsonl | 包含训练数据集，用于模型的训练过程。数据集经过标注，包含了事件触发词及其对应的事件类型标签。 Contains the training dataset used for the model's training process. The dataset is annotated, including event trigger words and their corresponding event type labels. |
| valid.jsonl | 包含验证数据集，用于在训练过程中对模型进行验证和调整。数据的标注与训练集一致。 Contains the validation dataset used to validate and adjust the model during training. The annotations are consistent with the training set. |
| test.jsonl | 包含测试数据集，用于在模型训练完成后对其性能进行评估。标注方式与训练集和验证集一致。 Contains the test dataset used to evaluate the model's performance after training is complete. The annotation method is consistent with the training and validation sets. |
| all.jsonl | 包含整个数据集的所有数据，包括训练集、验证集和测试集的所有内容，适用于全面分析或模型训练时的整体数据使用。 Contains all the data from the entire dataset, including the training, validation, and test sets, suitable for comprehensive analysis or using the entire data during model training. |
| doc2id.jsonl | 该文件为文档和对应ID的映射关系，帮助标注或训练时快速定位和检索相关文本。 This file maps documents to their corresponding IDs, aiding in quickly locating and retrieving relevant texts during annotation or training. |
| label.jsonl | 该文件定义了数据集中所有可能的事件标签，包含标签的详细信息和分类。 This file defines all possible event labels in the dataset, including detailed information and classification of the labels. |
| doccano_formatted_CHED_dataset.jsonl | CHED数据集的所有古文篇章级标注数据，可以导入doccano标注平台。 Contains all the ancient Chinese literary passage-level annotation data from the CHED dataset, ready for import into the Doccano annotation platform. |
数据集格式说明 | Dataset Format Description
句子级：
{
"sen_id": 1986, # 句子序号 | Sentence ID
"doc_id": 32029, # 文档编码 | Document ID
"text": "寻以本官行梁州刺史。", # 文本内容 | Text Content
"events": [ # 事件 | Events
{
"id": 3290, # 标签序号 | Label ID
"trigger": "行", # 触发词 | Trigger Word
"label": "职位-官位-任职", # 事件标签 | Event Label
"start_offset": 4, # 触发词在“text”中的起始索引位置 | Start Offset of Trigger in Text
"end_offset": 5 # 触发词在“text”中的终止索引位置 | End Offset of Trigger in Text
}
]
}篇章级：
{
"doc_id": "24明史卷一百二十三列传第十一方国珍", # 文档编码 | Document ID
"total_sentences": 93, # 总句子数 | Total Number of Sentences
"text": "方国珍，黄岩人。长身黑面，体白如瓠，力逐奔马。...", # 文本内容 | Text Content
"events": [ # 事件列表 | Events List
{
"id": 6704, # 事件序号 | Event ID
"trigger": "贩", # 触发词 | Trigger Word
"label": "经济-买卖-卖出", # 事件标签 | Event Label
"start_offset": 25, # 触发词在“text”中的起始索引位置 | Start Offset of Trigger in Text
"end_offset": 26 # 触发词在“text”中的终止索引位置 | End Offset of Trigger in Text
},
{
"id": 6705, # 事件序号 | Event ID
"trigger": "捕", # 触发词 | Trigger Word
"label": "法律-逮捕", # 事件标签 | Event Label
"start_offset": 53, # 触发词在“text”中的起始索引位置 | Start Offset of Trigger in Text
"end_offset": 54 # 触发词在“text”中的终止索引位置 | End Offset of Trigger in Text
}
# 其他事件按此格式依次列出...
],
"relations": [] # 事件间关系列表 | Relations List (为空时表示无关系)
}论文引用 | Citation
如果您在研究中使用了该数据集，请引用以下论文：
If you use this dataset in your research, please cite the following paper:
@inproceedings{congcong-etal-2023-ched,
title = "{CHED}: A Cross-Historical Dataset with a Logical Event Schema for Classical {C}hinese Event Detection",
author = "Congcong, Wei and
Zhenbing, Feng and
Shutan, Huang and
Wei, Li and
Yanqiu, Shao",
editor = "Sun, Maosong and
Qin, Bing and
Qiu, Xipeng and
Jiang, Jing and
Han, Xianpei",
booktitle = "Proceedings of the 22nd Chinese National Conference on Computational Linguistics",
month = aug,
year = "2023",
address = "Harbin, China",
publisher = "Chinese Information Processing Society of China",
url = "https://aclanthology.org/2023.ccl-1.74",
pages = "875--888",
abstract = "{``}Event detection (ED) is a crucial area of natural language processing that automates the extrac-tion of specific event types from large-scale text, and studying historical ED in classical Chinesetexts helps preserve and inherit historical and cultural heritage by extracting valuable informa-tion. However, classical Chinese language characteristics, such as ambiguous word classes andcomplex semantics, have posed challenges and led to a lack of datasets and limited research onevent schema construction. In addition, large-scale datasets in English and modern Chinese arenot directly applicable to historical ED in classical Chinese. To address these issues, we con-structed a logical event schema for classical Chinese historical texts and annotated the resultingdataset, which is called classical Chinese Historical Event Dataset (CHED). The main challengesin our work on classical Chinese historical ED are accurately identifying and classifying eventswithin cultural and linguistic contexts and addressing ambiguity resulting from multiple mean-ings of words in historical texts. Therefore, we have developed a set of annotation guidelinesand provided annotators with an objective reference translation. The average Kappa coefficientafter multiple cross-validation is 68.49{\%}, indicating high quality and consistency. We conductedvarious tasks and comparative experiments on established baseline models for historical ED inclassical Chinese. The results showed that BERT+CRF had the best performance on sequencelabeling task, with an f1-score of 76.10{\%}, indicating potential for further improvement. 1Introduction{''}",
language = "English",
}点击这里查看论文 | Click here to view the paper
数据分析 | Data Analysis
-
数据大小 | Data Size
我们构建的古汉语历史检测数据集，包含来自《二十四史》中61篇文档，61个历史人物。数据集总计包含13,159个句子和236,842个汉字。其中，有8,122个带有事件标签的句子，共计145,973个汉字，并且总共有14,154个标签。
The dataset we constructed contains 61 documents from the "Twenty-Four Histories" and covers 61 historical figures. It includes a total of 13,159 sentences and 236,842 Chinese characters. Among these, 8,122 sentences are annotated with event labels, totaling 145,973 characters, and there are 14,154 labels in total.
-
事件逻辑体系构建 | Event Logical Schema Construction
我们构建了一个完整、具有逻辑的古文历史事件类型本体。古文历史事件类型之间是有层次关系的，事件大类之间也是具有联系的，这形成了一个严密的逻辑体系，包括9大类和67小类的古文历史事件类型。
We constructed a comprehensive and logical schema for classical Chinese historical events. The event types are hierarchical and interconnected, forming a coherent logical system, which includes 9 major categories and 67 subcategories of classical Chinese historical event types.
| 一级类别 | 二级类别 | 三级类别 | 四级类别 |
|---|---|---|---|
| 人生 | 出生 | - | - |
| 人生 | 结婚 | - | - |
| 人生 | 生病 | - | - |
| 人生 | 受伤 | - | - |
| 人生 | 死亡 | - | - |
| 人生 | 安葬 | - | - |
| 人生 | 分封 | - | - |
| 人生 | 立谥 | - | - |
| 职位 | 帝位 | 确立 | - |
| 职位 | 帝位 | 即位 | - |
| 职位 | 帝位 | 继位 | - |
| 职位 | 官位 | 任职 | - |
| 职位 | 官位 | 免职 | - |
| 职位 | 官位 | 升职 | - |
| 职位 | 官位 | 降职 | - |
| 职位 | 官位 | 赏赐 | - |
| 交流 | 国家交流 | 出使 | - |
| 交流 | 国家交流 | 盟约 | - |
| 交流 | 国家交流 | 朝贡 | - |
| 交流 | 个人交流 | 见面 | - |
| 交流 | 个人交流 | 商议 | - |
| 交流 | 个人交流 | 询问 | - |
| 交流 | 个人交流 | 诏令 | 命令 |
| 交流 | 个人交流 | 诏令 | 下诏 |
| 交流 | 个人交流 | 进言 | - |
| 交流 | 个人交流 | 派遣 | - |
| 祭祀 | 祭天神 | - | - |
| 祭祀 | 祭地祇 | - | - |
| 祭祀 | 祭人鬼 | - | - |
| 移动 | 到达 | - | - |
| 移动 | 停留 | - | - |
| 移动 | 离开 | - | - |
| 军事 | 备战 | 出兵 | - |
| 军事 | 备战 | 驻扎 | - |
| 军事 | 作战 | 攻击 | 征伐 |
| 军事 | 作战 | 攻击 | 侵袭 |
| 军事 | 作战 | 防守 | - |
| 军事 | 作战 | 谋反 | - |
| 军事 | 作战 | 杀害 | - |
| 军事 | 作战 | 救援 | - |
| 军事 | 作战 | 撤退 | - |
| 军事 | 作战 | 俘虏 | - |
| 军事 | 停战 | 战胜 | - |
| 军事 | 停战 | 战败 | - |
| 军事 | 停战 | 投降 | - |
| 经济 | 赋税 | 征税 | - |
| 经济 | 赋税 | 减税 | - |
| 经济 | 赋税 | 免税 | - |
| 经济 | 买卖 | 买入 | - |
| 经济 | 买卖 | 卖出 | - |
| 经济 | 生产 | 耕种 | - |
| 经济 | 生产 | 打猎 | - |
| 经济 | 生产 | 建造 | - |
| 法律 | 制法 | - | - |
| 法律 | 废法 | - | - |
| 法律 | 犯罪 | - | - |
| 法律 | 逮捕 | - | - |
| 法律 | 赦免 | - | - |
| 法律 | 处死 | - | - |
| 法律 | 流放 | - | - |
| 自然 | 灾害 | 饥荒 | - |
| 自然 | 灾害 | 地震 | - |
| 自然 | 灾害 | 水灾 | - |
| 自然 | 灾害 | 蝗灾 | - |
| 自然 | 天象 | 日食 | - |
| 自然 | 天象 | 流星 | - |
| 自然 | 天气 | - | - |
许可证 | License
本数据集使用 CC BY-NC 4.0 许可证进行分发。您可以自由使用和分享数据集，但需注明出处，并且仅限于非商业用途。
This dataset is distributed under the CC BY-NC 4.0 license. You may freely use and share the dataset, but you must provide proper attribution and use it only for non-commercial purposes.
使用条款 | Terms of Use
-
数据用途: 数据集仅用于学术研究和教育目的，不得用于商业用途。 Data Usage: The dataset is intended for academic research and educational purposes only, and must not be used for commercial purposes.
-
数据保护: 使用数据时，请确保遵循相关数据保护法规和伦理规范。 Data Protection: When using the data, please ensure compliance with relevant data protection regulations and ethical standards.
-
引用要求: 使用数据集进行研究或发布结果时，请确保正确引用数据集和相关论文。 Citation Requirements: When using the dataset for research or publishing results, ensure proper citation of the dataset and relevant papers.