Querying Commonsense Knowledge

How to query commonsense knowledge from an Agent, Ontology and Relation oriented Commonsense Knowledge Base (AORKB) is presented. During querying, natural language queries are first transformed into descriptive queries in a kind of language-Agent and Ontology oriented Descriptive Query Language (AODQL) defined by us. Then, descriptive queries are translated into Structured Query Language (SQL) queries. After querying, query results in relation are converted into natural language by the natural language generating module. All tests show that the new method of querying is superior to the method of querying knowledge from agents and ontologies directly in terms of querying time. And up to 80% reduction in querying time is achieved.