We propose development of an adaptive, personalizable, information management tool, which can be configured and trained by an individual biologist to most effectively exploit the particular knowledge bases and document collections that are most useful for him or her. The proposed tool represents a novel approach for monitoring scientific progress in biology, which has become a formidable task. We will exploit recent advances in machine learning and database systems to develop a useful approximation to a personalized biological knowledge base f.i.i.e., single information resource that would include all the knowledge sources on which a biologist relies. More specifically, we propose a scheme for loosely integrating both structured information and unstructured text, and then querying the integrated information using easily-formulated similarity queries. The system will also learn from every episode in which a biologist seeks information. The research team on this project includes a computer scientist and two biologists. The proposed work will make systems for monitoring scientific progress in biology more effective. This will make biologists, clinicians and medical researchers better able to track advances in the biomedical literature that are relevant to their work.
| Talkowski, Michael E; McCann, Kathleen L; Chen, Michael et al. (2010) Fine-mapping reveals novel alternative splicing of the dopamine transporter. Am J Med Genet B Neuropsychiatr Genet 153B:1434-47 |
| Arnold, Andrew; Cohen, William W (2009) Information Extraction as Link Prediction: Using Curated Citation Networks to Improve Gene Detection. WASA 5682:541-550 |
