1 Introduction to the Algorithm
The Knowledge Graph is an important branch of artificial intelligence technology, proposed by Google in 2012. It is a structured semantic knowledge base used to describe concepts in the physical world and their interrelationships in symbolic form. Its basic unit consists of “entity-relation-entity” triples, as well as entity and its associated attribute-value pairs. Entities are interconnected through relations, forming a web-like structure of knowledge.
The Knowledge Graph can be divided into general knowledge graphs and domain-specific knowledge graphs based on functionality and application scenarios. General knowledge graphs target general fields, emphasizing the breadth of knowledge, usually in the form of structured encyclopedic knowledge, aimed primarily at ordinary users. In contrast, domain-specific knowledge graphs focus on a specific field, emphasizing the depth of knowledge, and are usually constructed based on industry-specific databases, targeting professionals within the industry and potential insiders.
The origin of the Knowledge Graph can be traced back to the 1960s. In the early development of artificial intelligence, there were two main branches: one is the symbolic approach, which focuses on simulating human cognition and studying how to represent knowledge in the human brain using computer symbols to mimic human thinking and reasoning processes; the other is the connectionist approach, which emphasizes simulating the physiological structure of the human brain, leading to the development of artificial neural networks. At this time, Semantic Networks were proposed as a method of knowledge representation, mainly used in the field of natural language understanding.
2 Algorithm Implementation
To establish a Knowledge Graph, the first step is to obtain data, which serves as the source of knowledge. This data can come from tables, text, databases, etc. Based on the type of data, it can be classified into structured data, unstructured data, and semi-structured data.
When data from different sources is obtained, it is necessary to perform knowledge fusion, which involves merging entities that represent the same concept and combining data from multiple sources into a single dataset. This results in the final dataset, upon which the corresponding Knowledge Graph can be constructed.
The Knowledge Graph can acquire new knowledge through techniques such as knowledge reasoning, thus continuously improving the existing Knowledge Graph.
3 Applications
The Knowledge Graph is a structured, visual semantic database and a branch of knowledge engineering, playing a crucial role in the field of artificial intelligence. The underlying logic of search engines we use daily and intelligent recommendations on e-commerce platforms all utilize Knowledge Graph technology. Over the past decade, research on traditional Chinese medicine (TCM) knowledge graphs has developed well, with an increasing number of related research papers published in China since 2012, making it a hot research area in TCM informatics.
In the clinical application research of TCM knowledge graphs, the focus is primarily on three aspects: visualization of clinical knowledge context, machine question-answering (medical knowledge generation), and clinical decision support. For example, in the research on establishing a Knowledge Graph for treating diabetic peripheral neuropathy (DPN), researchers constructed a TCM treatment knowledge graph based on a large amount of literature and case data, using the “disease-syndrome-treatment-prescription-drug” knowledge framework and conceptual logical relationships. From syndrome element analysis and syndrome composition to treatment method selection, a complete chain of knowledge structure for syndrome differentiation and treatment of DPN was formed, supporting queries and discovery of multiple nodes such as TCM treatment elements, syndrome types, treatment methods, and the relationships between these nodes. This not only facilitates the explicitization of implicit knowledge in TCM literature related to DPN treatment but also provides new ideas and methods for clinical and research work.
4 Conclusion
The TCM knowledge graph is currently in a rapid development stage, with various research results emerging continuously and large-scale knowledge platforms being constructed. However, challenges such as the lack of high-quality datasets, scarcity of talent resources, numerous interference factors in knowledge fusion, insufficient data sharing, limited generalization ability of graph construction and analysis algorithms, and lack of unified evaluation metrics remain.
The application of artificial intelligence and big data technology in the field of TCM remains a hot research direction for the future, mainly including improving data quality, enhancing knowledge association and mining algorithms, optimizing and applying knowledge graph visualization and interaction methods, etc. There is still vast potential for research and application of TCM knowledge graphs, requiring further study and exploration.
[1] Zhao Hanqing. Review of Research on TCM Knowledge Graphs [J]. China Digital Medicine, 2023, 18(08): 102-107.
[2] Chai Jiaqi, Tan Yumeng, Xiang Xinghua, et al. Construction of Knowledge Graph Based on Literature Data: A Case Study of TCM Treatment for Diabetic Peripheral Neuropathy [J/OL]. Chinese Journal of Experimental Traditional Chinese Medicine: 1-8 [2023-10-10].
[3] Zhihu Column. “Introduction to Knowledge Graph: Understanding Knowledge Graphs”. Accessed on October 10, 2023. https://zhuanlan.zhihu.com/p/396516565.
Recommended Reading:
Classification and Regression Trees (CART) – An Advanced Upgrade of Decision Trees
K-Nearest Neighbors (KNN) – “Birds of a Feather Flock Together”
Mining and Analyzing Professor Huang Li’s Medication Patterns for Treating Young Hypertension Based on Ancient and Modern Medical Cases Cloud Platform
Decision Trees – More Than Just Decisions
Ancient and Modern Medical Cases Cloud Platform
Providing Retrieval Services for Over 500,000 Ancient and Modern Medical Cases
Supports Manual, Voice, OCR, and Batch Structured Entry of Medical Cases
Designs Nine Analytical Modules, Close to Clinical Practical Needs
Supports Collaborative Analysis of Massive Medical Cases and Personal Medical Cases
EDC TCM Research Case Collection System
Supports Multi-center, Online Random Grouping, Data Entry
SDV, Audit Trail, SMS Reminders, Data Statistics
Analysis and other functions
Supports Customized Form Design
Users can log in to the website: https://www.yiankb.com/edc
Free Trial!
Institute of Information Research, China Academy of Chinese Medical Sciences
Smart R&D Center for TCM Health
Big Data R&D Department
Phone: 010-64089619
13522583261
QQ: 2778196938
https://www.yiankb.com
