Data Fusion - Summary
Data Fusion - integrating complex data from multiple sources
SUMMARY
We have seen that complex data correlation systems make high demands on the underlying DBMS. They need:
- High Performance with complex data – algorithms need the data at RAM speeds, not disk speeds. Data correlation systems may need to deal with batches of data or continuous streams from multiple sources.
- Scalability - in both data volume and the number of concurrent users (or threads).
- Reliability – many data correlation systems are deployed in 7 x 24 environments.
- Data Distribution – data may have to be globally available. It may be cached at one or more geographic sites to avoid depending on access to a central location.
- Interoperability – data may be captured and processed on differ ent kinds of equipment. It may be processed in C++ and accessed via tools built with Java; or generated on an RTOS, processed on a UNIX/Linux box and viewed on a Windows PC.
- Flexibility – the types of data and the relationships between data instances may need to be dynamically configurable by domain experts or end users.
- Additional functionality at a very low cost.
- Faster time to deployment because of the elimination of mapping code and Objectivity/DB’s powerful object modeling and manipulation features.
- Lower cost of ownership because it is easier to maintain, upgrade and expand deployed applications and systems.
- Dynamic adaption to changing requirements.

