Extract Medical Interpretation Based on Shallow Syntactic Analysis

Knowledge extraction provides the potential for producing high quality representation of the document and can help storing, retrieving, sharing, and management of explicit biomedical knowledge. This article addresses the task of mining binary relationships between concepts from biomedical literature for scientific discovery from medical literature. The UMLS has defined the domain entities and the internal relation network between them. The expert is then guided in construct the relation template by define the syntactic and semantic constraints with the help of the pre-established domain ontology. We present the technologies in finding the possible knowledge from plain text based on some grammar analysis. Medical concepts with the similar role are grouped as a union to improve the recall and decrease the useless calculate. The medical relations are extracted according the syntactic and semantic mapping. Experiment show that the biomedical extraction system get better performance.