Abstract
We present our approach to the problem of optimizing compression choices in the context of the XQueC compressed XML database system. In XQueC, data items are aggregated into containers, which are further grouped to be compressed together. This way, XQueC is able to exploit data commonalities and to perform query evaluation in the compressed domain, with the aim of improving both compression and querying performance. However, different compression algorithms have different performance and support different sets of operations in the compressed domain. Therefore, choosing how to group containers and which compression algorithm to apply to each group is a challenging issue. We address this problem through an appropriate cost model and a suitable blend of heuristics which, based on a given query workload, are capable of driving appropriate compression choices.
| Original language | English |
|---|---|
| Journal | Dagstuhl Seminar Proceedings |
| Volume | 8261 |
| Publication status | Published - 1 Jan 2008 |
| Externally published | Yes |
| Event | Structure-Based Compression of Complex Massive Data 2008 - Wadern, Germany Duration: 22 Jun 2008 → 27 Jun 2008 |