GDB: A tool to build deductive rules using a fuzzy relational database with scientific data

Abstract

This paper presents GDB, a system to build, debug and test rules in a Fuzzy Relational Deductive Database System (FRDDS) in a visual way avoiding the associated syntax which could be quite complex for a non-expert user. The rules will be used to compute a Measure of Quality (MoQ) of the scientific data generated by the GIADA instrument (ROSETTA space mission). The MoQ will be defined and assigned taking into account the knowledge of the engineers who built the instrument and the formal documentation. Together with this MoQ, the tool will provide a semantic explanation about how the measure has been assigned. With GDB is possible to work with imprecise data, to store knowledge using rules and also to deduce new information. The MoQ and the explanation will provide to the scientists and engineers a better understanding about the instrument behavior and, therefore, about physical phenomena under study.