PubMedCentral's Search Engine Adopts UI Changes Consistent with Recommendations from BioText Search Interface Research.
Several Biotext research papers ( Bioinformatics and PLos ONE) find evidence for the efficacy of showing figures in bioscience search results and allowing search over figure captions. An NLM Technical Bulletin from July 2011 describes adoption of these features for PubMedCentral publications which have the necessarily open intellectual property rights. (Perhaps the BioText talk at NLM in spring 2009 had some impact? We hope so!)
When the project began, new methods and tools were needed to improve how bioscience researchers search for and synthesize information from textual descriptions of bioscience research. This project built a flexible, efficient, platform-independent database system infrastructure specifically geared towards supporting the advanced and particular search needs of bioscience researchers. It used this infrastructure to support the development and deployment of statistical approaches to natural language processing, which was used to identify entities and relations between them in bioscience texts.
This project also worked with bioscience researchers to develop intuitive, appealing, interfaces for using these facilities to perform efficient and effective searches. The resulting BioText Search system supported new ways of asking scientific questions of the underlying databases.
Although the BioText Search Engine has been retired, that link shows the original homepage and screenshots of the results of some sample queries.