A TCM Diagnosis System Based on Textbook Information Extraction

The digitization of books can not only provide a electronic way to read but also make it possible to establish information systems via information extraction and correlation. This paper proposed a new approach that build up a traditional Chinese medicine (TCM) diagnosis system by extracting information from textbooks. First of all, we design a semantic ontology which can achieve targeted and precise extraction by combining it with SVM classification and regular expression match. After the popularization of the ontology, data and information in the database can be correlated automatically. Moreover, aiming at the domain of TCM textbooks, a new symptom identification algorithm is adopted. Eventually, a structural knowledge database is constructed and the experiment system shows that our method can be useful to provide new services for digital library.