The Design and Implementation of an Artifically Intelligent Personal Assistant Using Multiple Ontologies and Neural Ttoken Lists

We describe the design of an Artificially Intelligent Personal Assistant (AIPA) system whose theory is based on concepts from the artificial neural networks, triple-based knowledge representation that is used for both ontologies and their binary relationships, fast access via hash functions, and Class Algebra. The purpose of research and implementation is to study and test the effects of sharing multiple ontologies and Neural Token Lists (NTL), and to observe a Knowledge Sharing Connection (KSC) on the Internet, so that different knowledge of Closed Domains can be shared among humans and computer ontologies in order to significantly increase the knowledge content of the AIPA. KeywordsIntelligent Assistant、QA system、Semantic Network、Ontology、Machine Learning.