Design and Implementation of Mining Indirect Connections Between Medical Complementary Literatures Based on UMLS
基于 UMLS医学本体的挖掘文献间潜在联系的设计与实现

Following Swanson's knowledge discovery method from non-interactive literatures,the authors propose a literature-based discovery method by applying NLP techniques to finding biomedical Unified Medical Language System(UMLS) concepts.And then the authors introduce semantic filter system and develop a non-interactive medicial literature-based mining tool,using Swanson's discovery of connecting migraine with a magnesium deficiency to validate.Experiments show that this method can recur Swanson's research results very well and find more worthful medi-words.