Georeferencing Flickr photos using language models at different levels of granularity: An evidence based approach

摘要: The topic of automatically assigning geographic coordinates to Web 2.0 resources based on their tags has recently gained considerable attention. However, the that are produced by automated techniques necessarily variable, since not all described sufficiently descriptive. Thus there is a need for adaptive assign locations photos at right level granularity, or, in some cases, even refrain from making any estimations regarding location all. To this end, we consider idea training language models different levels and combining evidence provided these using Dempster Shafer's theory evidence. We provide experimental results which clearly confirm increased spatial awareness thus allows us make better informed decisions, moreover increases overall accuracy individual models.