Download now Free registration required
This paper present an approach for knowledge representation based on neural network, named Semantic Neuro Knowledge Representation system (SNKR). The proposal system represents knowledge as objects and relation between them, like semantic net. Each neural network in the system response to an input object by classifying it to a specific relation, more than one object can share the same relation. (SNKR) system can represent complicated knowledge which has complex relations, like propositional networks, more easily than conventional methods. Back Propagation (BP) neural network learning method are used to build neural networks of (SNKR) System because of its simplicity & easy of obstruction.
- Format: PDF
- Size: 132 KB