Platypus: A Constraint-Based Reasoning System

Platypus is a constraint-based reasoning engine for synthesis, diagnosis and other recognition tasks. While its target applications are similar to those of many rule-based expert system shells, the architecture of the underlying reasoning engine is not. Platypus is part of our goal at Tektronix of defining an architecture for Smart Instruments, In this architecture, a model-based knowledge representation provides rules for constructing object-centered descriptions from the input data. The descriptions produced make explicit the entities recognized in the task domain, their identifying parameters and the semantic constraints that exist among the entities. Platypus is described using a synthesis task, the configuration of Tektronix 4300-series workstations.

• Type: Conference
• Year: 1989
• Where: IJCAI
• Authors: William S. Havens, Paul Stephen Rehfuss
• Keywords: Constraint-based Reasoning Engine, IJCAI 1989, IJCAI 2007, Platypus, Reasoning Engine