Enhancing Multimedia Data Fragmentation

Distributed multimedia applications have emerged at an increasing rate during the last decade in several domains (video conferencing, e-health, virtual meeting rooms, etc). This has created several new challenging problems related to data integration and fragmentation, user-oriented and adaptive interfaces, real time and network performances, etc. In this article, we focus on the problem of data(base) fragmentation, initially consisting of reducing irrelevant data accesses by grouping data frequently accessed together in dedicated segments, in a multimedia context. We mainly address the issue of query and predicate implication required in current fragmentation algorithms, and provide a formal approach to identify such implications, in order to partition multimedia data efficiently. It is worthy to note that our approach is capable of considering multimedia-based as well as semantic comparisons, based on a generalized notion of functional dependencies, which are called multimedia functional dependencies.