Tsinghua University Statistics Research Center and Guangdong Hong Kong, Macau Digital Digital Economic Research Institute jointly developed large open medical knowledge map

November 22 – 23, Shenzhen Futian District People’s Government, Shenzhen Futian District Science and Technology Innovation Bureau and the Guangdong-Hong Kong Macao Digate District Digital Economic Research Institute ("Idea") jointly organized the IDEA Conference. Director of the Idea Shouyuan, foreign academician of the US National Academy of Engineering, External Academician of the UK Royal Engineering, Double Professor Tsinghua University Double Professor Shen Yanyong released a number of prociation teams and Guangdong and Hong Kong, Macau Duwan District, Association A large-scale open medical knowledge map (Biomedical Informatics Ontology System, referred to as "BIOS").

Shenyang explains in the BIOS conference

Medical knowledge map is a special database consisting of biomedical concept name, concept classification, conceptual relationship, and corresponding ID system. It is used to support medical natural language processing, artificial intelligence modeling, and industry data exchange. It is medical data and One of the most important infrastructure in artificial intelligence is a strategic impact on industry development. The National Medical Library of the US National Health Center has developed and developed in 1986, the unified Medical Language System, referred to as "UMLS") is the most authoritative English medical knowledge map, which is the English national medical data technology and The development of the industry made an excellent contribution. For a long time, the lack of large-scale open medical knowledge maps in the Chinese field is one of the main constraints of my country’s medical big data and artificial intelligence industry.

Description of the number of entities and relationships included in the BIOS knowledge map

Medical naming entity identification demonstration based on BIOS knowledge map

In order to solve the lack of medical knowledge maps in the Chinese field, and further enhance the construction level of medical knowledge maps in the international scope, Yu Zun leads the team to conduct a five-year technical research, and has developed Chinese based on map segmentation and depth learning. Chinese Electronic medical records do not supervise multi-granular particless words and terminology extraction, knowledge determined by medical terms to quantization and regularization, high-throughput medical relationship, biomedical translation, etc. Cooperating with the Idea Research Institute in November 2020 to reach a cooperation. With the support of leading algorithms, powerful strength and oversized corpus data, only in just one year, the two teams started from the original underlying medical terminology, and established a new Chinese and English bilingual medical knowledge map with complete independent intellectual property rights. BIOS, its size is close to existing authoritative knowledge map UMLs, and forming a multi-point transcendence in content quality.

The BIOS is currently published (http://bios.idea.edu.cn). At the same time, adhering to the comprehensive improvement of the development of China’s medical big data and artificial intelligence industry, BIOS is in the near future to open full data downloads with CCBY-NC-ND protocol.

In the future, Tsinghua University Statistical Research Center will cooperate with the IDEA Research Institute and more domestic top hospitals, continuously expand the content of BIOS and improve their quality. This will not only make the basic improvement of medical high data and artificial intelligence industries, but also the radiation international and drive the global industry.