Imposing a Semantic Schema for the Detection of Potential Mistakes in Knowledge Resources

Society is becoming a complex socio-technical ecosystem requiring novel ICT solutions to provide the basic infrastructure for innovative services able to address key societal challenges and to assist people in their everyday activities. To tackle such complexity, there is a pressing need for very accurate, up-to-date and diversity-aware knowledge resources which can guarantee that results of automatic processing can be trusted enough for decision making processes. As the maintenance of such resources turns out to be very expensive, we argue that the only affordable way to address this is by complementing automatic with manual checks. This paper presents a methodology, based on the notion of semantic schema, which aims to minimize human intervention as it allows the automatic identification of potentially faulty parts of a knowledge resource which need manual checks. Our evaluation showed promising results.