A hybrid model approach to artificial intelligence

This paper establishes a learning paradigm for natural language. The paradigm is then effectively modeled by using a hybrid approach to Artificial Intelligence (AI). A synergy of established AI paradigms was found by examining various methodologies, and choosing aspects of each that can replicate the desired learning paradigm. The hybrid model postulated is a combination of object-oriented deductive database with elements of fuzzy logic, neural networks, and natural language understanding methodologies. A three-word, transitive verb-phrase, natural language understanding model was created, based on fuzzy values, and inference mechanisms that endow the model with deductive capabilities. The resulting model creates new phrases based on input of data, while queries return information that is relevant to the universe of the models matrix. The resulting output is based on a dynamic relationship between the static objects in the matrix and the fed input. The results showed that the mechanisms involved in this model can be effective, and realistically implemented.