Literature-based discovery has been much discussed in a variety of contexts including information retrieval, computer science, computational linguistics and bio-informatics, and research in the area is growing at an impressive pace. Unlike text mining, literature-based discovery seeks to uncover information that is implicit but not explicitly stated within a literature, or to find concepts that link two disparate literatures. An important body of work has developed around Don Swanson's pivotal contributions, originating with his 1986 paper on the plausibility of literature-based discovery. His first case of discovery in disconnected literatures in biomedicine, suggesting the therapeutic effects of fish oil on Raynaud's Syndrome, was also published that year. Two years later, Swanson reported another literature-based hypothesis intimating the involvement of magnesium insufficiency in migraine headaches. Both predictions were subsequently confirmed.
Literature-based discovery publications have been produced in biomedical domains by the collaborative team of Swanson and Smalheiser; Weeber and Vos; Hristovski; and Gordon and colleagues; and several other groups are working actively in this area. This panel brings together Don Swanson and other active text retrieval researchers to present current trends, discuss roadblocks, and forecast new opportunities in literature-based discovery. The last segment of the panel will be devoted to discussion among the group as a whole, with additional discussants being invited from the larger ASIS&T research community.
1. Don Swanson, Univ. of Chicago, keynote speaker. 2. Padmini Srinivasan, University of Iowa 3. Marti Hearst, University of California at Berkeley 4. Neil R. Smalheiser, University of Illinois at Chicago
Carole Palmer, Univ. of Illinois at Champaign-Urbana, will serve as moderator.