ISI's Scientific Knowledge Engineering ('SKE') group combines ontology/knowledge engineering, artificial intelligence, the semantic web, machine reading and science informatics techniques to support the development of next-generation scientific informatics infrastructure. We leverage general research techniques across information-intensive disciplines, focussing primarily on specialized subfields of biomedical informatics, but also generalizing to other fields. The group is lead by Gully Burns, a Research Lead at the Information Sciences Institute.
The SKE group currently is pursuing automatic machine reading systems for scientific evidence, unsupervised modeling of complex research literature, modeling experimental design to better understand the structure of scientific data, automated scientifically-aware web-agents, application of information integration and semantic web to biomedical databases and pedadogical systems to support improved technical knowledge acquisition. Our work focuses on solving real-world problems to bridge research and relevant applications by building and applying tools for working scientists.
Our work has included development of:
- The BioScholar and NeuroScholar systems as prototypes of general purpose scientific knowledge management systems
- The Knowledge Engineering from Experimental Design (KEfED) methodology as a practical method for developing ontologies for complex, experimental biomedical data
- The Scientific Discourse Tagger (SciDT) system (in collaboration with Ed Hovy's group at CMU)
- The TechKnAcq tool for extracting reading lists from copora of scientific documents
- The NeuARt neuroanatomical data viewer
- The NIHMaps topic mapping system neuroanatomical data viewer
The SciKnowEngine 'swirl' logo represents the Cycle of Scientific Investigation as a group of combined threads that resembles a breaking ocean wave. The color scheme is derived from Van Gogh's Starry Night.