Özet:
One of the most important components of e-commerce business is electronic product catalog. In order to provide higher customer satisfaction, the accessibility of products or product groups should be quick and easy. The research done in the literature on efficient storage structure for XML is mostly on generic XML documents. In this work, we propose a categorized storage structure for text-based natively stored XML electronic product catalogs. Categorization method allows us to store XML-based product catalogs in an efficient way. Beside the storage of data, filtering of data becomes a vital process. Skyline is one way of doing filtering on data sets. In the literature on Skyline, many algorithms have been developed, but still there exists no algorithm that finds the Skyline of an XML-based dataset directly. As our second contribution, we keep an index holding the Skyline sets of frequently asked Skyline queries. So, in case of repetition of that queries our index makes the execution of any Skyline algorithm unnecessary. We present the performance comparison of our categorized storage structure with another approach for XML-based product catalog which considers their specific needs and properties. We also present the performance test of our Skyline index for various cardinalities.|Keywords: Electronic product catalogs, Semi-structured XML document, Categorization, Efficient query processing, Skyline operator, Indexing