Introduction
In the past two years, fields such as knowledge graphs, question answering (QA) systems, and chatbots have become increasingly popular. A knowledge graph is a generalized concept; in my view, any machine learning content involving the construction, retrieval, and utilization of knowledge bases can be considered part of a knowledge graph. Of course, this is not a formal definition, just a personal intuition.
Readers working on knowledge graphs know that triples are a method for structuring knowledge and are an important component of knowledge-based QA systems. In the English domain, there are already several large open-source triple corpora. Obviously, Chinese currently lacks such shared corpora (even if someone has crawled them, they are often kept private). Some time ago, I wrote a crawler for Baidu Baike and ran it for a while, capturing several million entries. Many of these entries contain structured information that, when extracted directly, forms valid "triples" suitable for building knowledge graphs. The triple corpus shared in this article is derived from this process, totaling 25 million triples.
Preview
The structure of a triple is (Entity, Attribute, Value). A partial preview is as follows:
Science, Includes, Fields such as nature, society, and thought
Science, Foreign Name, science
Science, Pinyin, kē xué
Science, Chinese Name, Science
Science, Explanation, Application and practice of discovered and accumulated truths
Grammar, Foreign Name, syntactics
Grammar, Chinese Name, Grammar
Physical Cosmology, Object, Large-scale structure and cosmic formation
Physical Cosmology, Time, Twentieth century
Physical Cosmology, Affiliation, Astrophysics
Physical Cosmology, Chinese Name, Physical Cosmology
Cao Yu, Birthplace, Tianjin
Cao Yu, Alma Mater, Tsinghua University
Cao Yu, Date of Death, December 13, 1996 (Year of Bingzi)
Cao Yu, Chinese Name, Wan Jiabao
The file is in CSV format with UTF-8 encoding. The total number of triples is 25,454,710, and the total number of entities is 4,695,579.
Since these were directly extracted from data manually edited by users, they tend toward natural language descriptions. This leads to some inconsistencies in descriptions: attributes with the same meaning may have multiple different descriptions. For example, "Birthplace" and "Born in" both represent the place of birth, while "Foreign Name" and "English Name" both represent the English name, etc.
Download
In the spirit of open source, these resources are shared for free. However, considering the hardship of crawling, please cite the URL of this article when reposting or quoting: http://kexue.fm/archives/4359/. Thank you.
Download Address:
Link: https://pan.baidu.com/s/1mkcKP2C
Password: uajy
Please include the original address when reposting: https://kexue.fm/archives/4359
For more detailed reposting matters, please refer to: Scientific Space FAQ