Fusion of meteorological and air quality data extracted from the web for personalized environmental information services

摘要: There is a large amount of meteorological and air quality data available online. Often, different sources provide deviating even contradicting for the same geographical area time. This implies that users need to evaluate relative reliability information then trust one sources. We present novel fusion method merges from given time, ensuring best quality. The unique combination land-use regression techniques, statistical modelling well-known algorithm. show experiments where fused temperature forecast outperforms individual forecasts several providers. Also, we demonstrate local hourly NO2 concentration can be estimated accurately with our while more conventional extrapolation falls short. forms part prototype web-based service PESCaDO, designed cater personalized environmental users. introduce information.We time.With presented pollutants assessed in varying urban environments.We system called AirMerge, which converts image-based maps into numerical fusion.