Output details
11 - Computer Science and Informatics
University of Cambridge
Natural language processing in aid of FlyBase curators.
<22> The paper describes a large system developed on a BBSRC
e-Science grant which is shown experimentally to improve FlyBase
curator productivity by 20%. It led directly to the formation of
Camtology Ltd., and the university also obtained a STFC MiniPIPSS
knowledge transfer grant (Scalable and Robust Grid-based Text Mining
of Scientific Papers, 2008-9) to help commercialise the results. This
was one of the first biomedical information extraction systems to be
successfully evaluated in a realistic operational context and was
entirely developed using weakly-supervised machine learning methods
achieving state-of-the-art performance for several subtasks.