Backforneb: uma ferramenta para construção de sistemas especialistas diagnosticadores.

Abstract

This paper describes the design and implementation of a tool for developing general purpose diagnostic systems (BACKFORNEB). Knowledge is represented in the form of rules and the promissory propagation of facts in the network will be done through forward and backward chaining, using fuzzy inference. BACKFORNEB consists of three main modules: EDIT, RUN, and I / O OPERATIONS. The EDITAR module, called EDICON, helps the Knowledge Engineer build the Desired knowledge in an organized and structured manner. The module RUN, is the system's engine of inferences and will have a large interacting with the user, asking them questions, receiving answers that were requested or spontaneously provided, and finally presenting possible diagnoses. I / O OPERATIONS are necessary for good use of the tool. This work also describes the user guide of the user and the system implementation aspects.