Biomedical literature is a hive of valuable information on research topics like diseases, drug/treatment attributes, medical decisions, health effects, population data and epidemiology, and more. With advances in technology, there is a rapid growth in the amount of this literature - making it impossible for researchers and practitioners alone to exhaust all of this valuable information.

With KNIME Software, mining knowledge from text such as disease-related information can be automated. An analytics expert creates a workflow in KNIME Analytics Platform, which contains a model that learns disease names from a set of documents in the biomedical literature. The trained model is then deployed to the KNIME WebPortal via KNIME Server. Here, with the predetermined interaction points, researchers can interactively inspect the diseases that co-occur in the same documents and explore genetic information associated with these diseases.

StanfordNLP nodes in the KNIME Textprocessing Extension (within KNIME Analytics Platform) facilitate building and evaluating the model. This extension also offers nodes for analyzing the results - for example Term Co-Occurrence Counter to investigate co-occurring diseases. Networking Mining nodes make it possible to visualize and analyze results and KNIME Server makes these interactive results accessible to researchers and domain experts.