KEfED - Knowledge Engineering from Experimental Design

This is a project that is under development within the Biomedical Knowledge Engineering Group at ISI (BMKEG (at) ISI) to provide an instantiation of the Knowledge Engineering from Experimental Design (KEfED) model. It is a structured way of constructing 'observational assertions' based on statistical relationships from experiments. The model is general-purpose and forms a basis for reasoning over experimental data. The KEfED model is designed to provide a lightweight representation for scientific knowledge that is (a) generalizable (b) a suitable target for text-mining approaches (c) relatively semantically simple and (d) is based on the way that scientist plan experiments and should therefore be intuitively understandable to non-computational bench scientists.The basic idea of the KEfED model is that scientific observations tend to have a common design: there is a significant difference between measurements of some dependent variable under conditions specified by two (or more) values of some independent variable.Version Zero: Basic Idea - is the first attempt to seriously define and instantiate the KEfED model as software. This is done in Java as a simple Object-Oriented web application.Version 0.1: Online KEfED editing system + database. This is the application that permits the user to design a KEfED model. This is a preliminary demo prototype system. Here we provide basic design documents and instructions for the software's deployment and use.

Resource Type: 
Parent organization: 
Biomedical Informatics Research Network
Supporting agency: 
Michael J. Fox Foundation for Parkinsons Research NCRR NIH
Grant: 
R01-GM083871
PMID: 
1R01MH079068-01A2