Towards automatic extraction of research findings from the literature.
Dimitar Hristovski, Debra Revere, Paul F. Bugni, Sherrilynne S. Fuller, Carol Friedman, Thomas C. Rindflesch · PubMed · 2007
Most information extraction systems only process an article's title and abstract. However, a major source of research findings is an article's tables and figures. The aim of this study is to: (1) explore the efficacy of applying a hybrid information extraction system to the problem of identifying research findings in the scientific literature and (2) improve results by processing article title, abstract and table/figure text.