Knowledge extraction and mining in biomedical research using rule network model

Recent findings show that the quantity of published biomedical literature is increasing at a dramatic rate. Carrying out knowledge extraction from large amounts of research literature becomes a significant challenge. Here we introduce an automatic mechanism for processing such information and extracting meaningful medical knowledge from biomedical literature. Data mining and natural language processing (NLP) are applied in a novel model, called biomedical rule network model. Using this model, information and relationships among herbal materials and diseases, as well as the chemical constituents of herbs can be extracted automatically. Moreover, with the overlapping chemical constituents of herbs, alternative herbal materials can be discovered, and suggestions can be made to replace expensive treatment options with lower cost ones