Application of NER and Association Rules to Traditional Chinese Medicine Patent Mining

Abstract

Traditional Chinese Medicine (TCM) patents contain much valuable information, and this paper studies the application of named entity recognition (NER) and association rules for the TCM patent mining. A NER approach combining the method based on rules and bootstrapping is proposed to recognize TCM entities in a large number of TCM patents. Besides, to explore the law of medicine use, we use the Apriori algorithm to generate association rules between TCM entities. By analyzing these association rules, we verify the effectiveness of association rules mining for the TCM patent mining, which provides a reference for the TCM treatment and research.