Event-based retrieval from digital libraries containing streamed data

Abstract

Data streams are a very important source of information for both researchers and ordinary users. Data streams might be video or audio streams or might be a stream of sensor readings, satellite images, etc. Saving data streams in digital libraries add the advantages of using digital libraries from archival, preservation, administration, access control, etc. to this information. For this specific type of digital libraries, event based retrieval provides an alternative, yet very natural way of retrieving information. People tend to remember or search by a specific event that happened in the stream better than by the time at which this event happened. The authors describe the organization and services of a prototype digital library system that contains streamed data and enables event based retrieval from them. They present a preliminary formal definition of data streams and events. They also show some possible applications of this system and conclude by presenting the ongoing and future extensions to the research.