TCMKG: A Deep Learning Based Traditional Chinese Medicine Knowledge Graph Platform
Authors:
- 1 August 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 560-564
Abstract
As an effective and novel knowledge management technology, knowledge graph can provide a new way for the inheritance and development of traditional Chinese medicine (TCM). However, the construction of the knowledge graph of TCM is still mainly based on structured data at present. With the accumulation of literatures and electronic medical records, a large amount of knowledge is stored in unstructured texts which urgently needs to be extracted for learning. In this study, we extract TCM core concepts and build ontology layer by analyzing the process of TCM diagnosis and treatment. Then we use deep learning to extract entities and their relations for building TCM knowledge graph from unstructured data. Finally, we build an end-to-end platform TCMKG based on knowledge graph, which can provide functions such as knowledge retrieval, visualization and data management for helping the learning and sharing of TCM knowledge.Keywords
This publication has 15 references indexed in Scilit:
- Construction of Semantic-Based Traditional Chinese Medicine Prescription Knowledge Graph2018
- Knowledge Graph Construction Based on Judicial Data with Social Media2017
- Joint entity and relation extraction based on a hybrid neural networkNeurocomputing, 2017
- Knowledge graph for TCM health preservation: Design, construction, and applicationsArtificial Intelligence in Medicine, 2017
- A Local Detection Approach for Named Entity Recognition and Mention Detection2017
- Chinese enterprise knowledge graph construction based on Linked Data2015
- Ontology-Oriented Diagnostic System for Traditional Chinese Medicine Based on Relation RefinementComputational and Mathematical Methods in Medicine, 2013
- Bio2RDF: Towards a mashup to build bioinformatics knowledge systemsJournal of Biomedical Informatics, 2008
- Freebase2008
- Yago2007