Extensible Indexing for XML Databases

XML databases require index structures to accelerate queries on the content and structure of XML documents. Recently, a large number of XML index structures has been proposed, each of which better supports different queries using proprietary data structures and algorithms. To support query workloads of arbitrary applications, an XML database would consequently require a large number of different index structures. Instead of proposing another index structure, this paper applies the concepts of extensible indexing to XML. By extending and adapting existing index structures to various requirements, a small number of index structures are sufficient to support a wider range of queries than existing approaches. To process index structures, the paper proposes an index framework that selects, accesses and maintains arbitrary indices based on a generic index model. Extensible indexing enables XML databases to provide those indices that best match the query workload.