Domain-Aware Data Matching in Emergency Response Scenarios

点击翻译按钮获取中文摘要

Nowadays, there are a lot of situations where participants from different domains have to collaborate. For example, in emergency response (ER) scenarios different agencies have to work together in order to reestablish normality as fast as possible. Nevertheless, this cooperation is not easy because the agencies involved in such scenarios represent knowledge in different ways. Although there are applied some techniques in order to address this heterogeneity problem, these sometimes produce mismatches caused by ambiguity and different knowledge representation degree between agencies (specialisation). In this research we are going to use domain-specific terminologies, to tackle these problems, instantiating our approaches into ER scenarios.At this point, we are formalising terminologies from ER agencies. Concretly, we are taking terms from the UK Civil and Protection glossary, which cotains common terms used by ER organisations in the UK. The next step consist in integrating these specific terms into a lexical resource (WordNet) and using these new terms to carry out semantic matchings between different agencies' schemas. Thus, we are going to take advantage of domain information in order to takle ambiguity and specialisation.