A Queriable XML Compression using Inferred Data Types

HTML is mostly stored in native file systems instead of specialized repositories such as a database. Like HTML, XML, the standard for the exchange and the representation of data in the Internet, is mostly resident on native file systems. However. since XML data is irregular and verbose, the disk space and the network bandwidth are wasted compared to those of regularly structured data. To overcome this inefficiency of XML data, the research on the compression of XML data has been conducted. Among recently proposed XML compression techniques, some techniques do not support querying compressed data, while other techniques which support querying compressed data blindly encode data values using predefined encoding methods without considering the types of data values which necessitates partial decompression for processing range queries. As a result, the query performance on compressed XML data is degraded. Thus, this research proposes an XML compression technique which supports direct and efficient evaluations of queries on compressed XML data. This XML compression technique adopts an encoding method, called dictionary encoding, to encode each tag of XML data and applies proper encoding methods for encoding data values according to the inferred types of data values. Also, through the implementation and the performance evaluation of the XML compression technique proposed in this research, it is shown that the implemented XML compressor efficiently compresses real-life XML data lets and achieves significant improvements on query performance for compressed XML data.