Abstract:

In many applications, data not only take the form of persistent relations. but also arrives in a real-time, continuous, time-varying and rapid way. Be  cause of these characteristics, many challenging issues are introduced in data processing. The purpose of this paper is to study the issues of data stre am modeling, and we present the approach of sliding window modeling using multiple time granularities, and characterize data streams by a hierarchical w indows model. And it also proves to be effective in solving the problem of historical data management of ad-hoc queries.

Key words: data stream model, multiple granularities, sliding window