A NLP Based Framework to Support Document Verification-as-a-Service

Many enterprise systems are document intensive that requires extensive manual verification in the form of maker and checker. However, a maker-checker based verification raises several challenges with respect to increase in cost and time of verification. Furthermore, any manual labor intensive verification is not free from human oversight and can lead to costly errors. Therefore, to alleviate the challenges arising out of human verification of document intensive systems, we propose a rule based framework that enables automatic verification of document based systems. The framework uses ontology based knowledge representation techniques along with appropriate natural language processing methods to extract operational rules from business documents and then use suitable reasoning engine for verification. The above framework is validated in the light of a real life case study namely International Trade that deals with several critical financial documents like Letter-of-Credit, Bill-of-Lading, Commercial Invoice etc.