Metamodeling to Ontology Conversion with Social Media Influence in Flood Management Domain

Abstract

Disaster is an unpredictable event unknowingly to human being when and where it will happen. Nevertheless, early precautions can be made by initiating Disaster Management (DM) to reduce losses and destructions that may arise. Based on the impact seen from recent events globally, DM has been a focus area constantly developing and evolving. Leveraging emerging technologies such as social media can improve actionable intelligence in Situational Awareness aspect. Among other meteorological events, flood has been the most common natural disaster by far, cumulating in total 43% out of all disaster phenomenon from the last 20 years. However, in DM perspective, flood is still informally represented to enable information exchange and comprehensive flood management activities. To address both issues, this study will further extend existing Disaster Management Metamodel (DMM) with social media concepts and designs Flood Management by applying metamodeling process with UML class diagram since it has been well known by domain experts. Still, class diagram unable to cater applications that relies on processing content of information instead of visualization only. This limitation is overcome by transforming the metamodeling source into ontology format which elevates the expressiveness, meaning, and delivering more structured information among human or software agents. The reconstructed design of DMM successfully identifies new concepts generally and for flood domain. After running the mapping algorithm, the generated ontology is validated with Protégé to ensure the markups are persistence with the latter.