The Biomedical Knowledge Engineering group at ISI develops practical, knowledge-centric systems for biomedical researchers. We focus on finding the strategic information bottlenecks in the current biomedical informatics architecture and work with top-level computer scientists to create well-engineered solutions and infrastructure.

Our work broadly follows the Basic Process Model (shown to the right).

The primary research literature provides the most comprehensive and authoritative source of information for biomedicine. It is the universal repository for individual research findings, their interpretations and their syntheses. The rate-determining step for the creation and population of many biomedical databases is process of biocuration. This requires experts (often at the level of Ph.D. workers) to read each article and enter their information into the relevant database. This is slow, difficult and expensive. We are developing computational solutions to accelerate the process of biocuration with Natural Language Processing techniques.

Scientists use formal, strict priniciples to construct scientific arguments based on carefully designing experiments that test hypotheses and search for . A large thrust of our work centers on developing a powerful new approach to represent the design of observations from scientific experiments that can support reasoning and (importantly) is immediately comprehensible to researchers.

Ultimately, we seek to transform the computational infrastructure supporting the scientific enterprise and specifically to accelerate methods of discovering and delivering improved medical treatment (i.e., 'translational research').

This is a very exciting and challenging field, requiring expertise from multiple fields and we are seeking partnerships with end users (both basic and medical scientists in academia and industry), government agencies, disease foundations, hospitals and the public.

If you have similar interests to us, or are interested in collaborating, please contact us.