Discovering disease-associated drugs using web crawl data

The purpose of research on biomedical literature-based discovery is to bring out new knowledge from the existing biomedical information. Beginning with Dr. Swanson's ABC model, many studies extended or applied the ABC model to find new associations between biomedical entities. While the methods applied to data have advanced, in most cases biomedical literature has been used for the text data. Assuming that web crawl data is helpful in studying literature-based discovery as well as biomedical literature which is the existing but rather limited data source, we discovered new disease-drug associations using web crawl data in addition to biomedical literature. We also analyzed how helpful the additional use of web crawl data is for biomedical literature mining. Literature-based discovery using web crawl data has its significance as a pioneering work utilizing new data.