Knowledge-based question answering with human plausible reasoning

This work explores the application of Human Plausible Reasoning (HPR) in building knowledge-based question answering systems. TeLQAS is an ontology-based domain-specific QA system for Telecommunications. We describe an attempt to use HPR as defined by Collins and Michalski as the core reasoning component of that system. The basic idea is to infer plausible answers for questions when exact answers are not available in the knowledgebase. A subset of available inferences in the theory has been implemented. The experiments conducted with Telecommunications knowledgebase show great potential and full implementation and larger experiments are under way.