SqueezeX: synthesis and compression of XML data

XML is emerging as the "universal" language for semi-structured data description/exchange, and new issues regarding the management of XML data, both in terms of performance and usability, are becoming critical. The application of knowledge-based synthesis and compression methods (i.e. derivation of synthetic views and lossless/lossy approximation of contents) can be extremely beneficial in this scenario. In this paper, we present SqueezeX, a system for the synthesis/compression of XML data based on a multidimensional/classification-based interpretation of their schema and on the application of semantic and type-oriented compression techniques.