Sensing Geographical Impact Factor of Multimedia News Events for Localized Retrieval and News Filtering

News materials are reports on events occurring in a given time and location. Looking at the influence of individual event, an event that has news reported worldwide is strategically more important than one that is only covered by local news agencies. In fact, news coverage of an event can accurately determine the event's importance and potential impact on the society. In this paper, we present a framework which extracts the latent impact factor of events from multimedia news resources by a geographical approach to support: (a) localized retrieval for end users; and (b) pre-screening of potential news elements that should be filtered for use by web monitoring agencies.