BioScholar consists of two curation tools for general use. The first tools shows a worked example of a knowledge-engineering web-application data that permits an end user to browse and query a knowledge base of observations and interpretations concerned with neural connectivity and tract-tracing experiments. A live demonstration as well as downloads of the application are available. The second is a domain-independent tool that permits users to curate their own models of scientific experiments and build small-scale data repositories for those models. This is the core functionality of the BioScholar system.
The BioScholar system uses a graphical interface to create experimental designs based on the experimental variables in the system. The design is then analyzed to construct a tabular input form based on the data flow. We call this methodology 'Knowledge Engineering from Experimental Design' or 'KEfED'. The approach is domain-independent but domain-specific modules reasoning can be constructed to generate interpretations from the observational data represented in the KEfED model.
This diagram shows an experimental design. The design illustrates the work and dataflow of an experiment. Individual elements of the design are then edited using the panels on the right side of the window. Based on the experimental design, BioScholar is able to generate tables for entering the data generated by the experiment. The data forms take care of tracing the data dependencies and creating a form for recording the necessary parameters for each measurement.
Domain specific intepretations are used to display and retrieve interpretations of the data from an experiement. In this example we use information from tract-tracing experiments to show connections between regions of the hippocampus in the rat brain. Each connection can be traced back to a particular experiment's data and linked to the scientific literature where that result was published.