MODEL-DRIVEN, ASSERTION-BASED GENERATION OF MULTIMEDIA WEATHER INFORMATION

As gridded forecasts are becoming a common source of weather data, eff orts are being made to automatically produce displays for various cate gories of decision makers such as emergency managers and road traffic officers. This paper proposes a new systematic approach to the generat ion of multimedia weather reports tailored to the specific needs of th e user. The process consists of two steps: extraction of assertions fr om gridded datasets, and their presentation in the form of maps, chart s, tables, and natural language text. The assertions organized by disc ourse relations relate weather elements to regions and periods of inte rest to the user. The user can explore the situation in a hypermedia m ode by descending from summary surveillance displays to more detailed exploratory ones. This approach provides a structure fora class of app lications dealing with the dissemination of weather information and su pports them with general methods of assertion extraction and media gen eration. The application-specific knowledge is localized in five ortho gonal and easily maintained models: territory, time, parameters, pheno mena, and events. A prototype system has demonstrated the viability of the concepts.