Wednesday, 20. March 2013 20.03.13 09:31 Age: 5 Jahre

Working Without Detours
Direct Data Access to Compressed Data

Compressed data made large – directly visible and immediately editable

Data compression – a standard procedure and a time-consuming one when handling the data: compressing, unzipping, searching, editing, recompressing, and starting all over again the next time. Used everywhere, requiring computation time over and over again. Not only for efficient data storage, but especially in a mobile environment as well. Here, data compression is used to keep the data traffic low and to go easy on the storage capacity. Compressed data is being used without cease – you cannot work with the data directly.

Standard data compression, such as ZIP, reaches a compression of about 30-60%. To achieve this, processing the compressed data always requires the decompression-editing-compression cycle, which means: time-consuming unzipping before modifications and time-consuming zipping afterwards. Even mobile devices constantly send and receive steadily increasing amounts of data. But only a limited extent of unzipping is possible because the internal memory is limited.

Using indexed reversible transformation (IRT), Prof. Dr. Böttcher, a member of the Electronic Commerce and Databases research group in the Institute of Computer Science at the University of Paderborn, has developed a method that not only enables processing the compressed data, but also performs a full text search efficiently in terms of memory and computation. In particular, it can be used especially to alter compressed data without complete decompression being necessary. When combined with XML schema definition subtraction (XSDS), the XML compression method that was also developed by the research group, the method can be applied not only to plain text files but also to the frequently used XML files.

The IRT and XSDS methods are perfect for helping devices with extremely limited resources, such as smartphones, reach a wider function scope even at low data transfer rates. IRT and XSDS provide the opportunity to reduce the data traffic and workload for the processor – and therefore time and energy as well: Processing compressed data faster saves time and storage space. At the same time, this increases the performance of the currently running application and also of the mobile devices altogether.

In general, many different kinds of applications, such as web browsers, social media, office tools, and financial transactions via SEPA (single euro payments area) can be improved. Without any problems it is possible to integrate IRT and XSDS technologies into existing solutions such as applications for smartphones and computers. This usually reduces the data volume by 90% to approximately 10% of the total size, which is smaller than by compression with gzip or bzip2, and also allows the data to be searched and even altered without any complete decompression.

Not only in the mobile sector does Prof. Dr. Böttcher foresee possible applications: The current versions of Microsoft Word and Excel also store their data as XML documents. When the IRT and XSDS methods are used here, not only are the memory requirements reduced, but a fast full-text search can also be performed on conventional computers. This is an attractive option – especially in light of the large amount of files that accumulate over time.

The IRT method is currently under application for a patent. In addition, the project partner PROvendis GmbH (the patent licensing agency for the universities in North Rhine-Westphalia) is offering licenses to interested companies.

Contact:
Prof. Dr. Stefan Böttcher
Electronic Commerce and Databases
Tel: 05251 60-6662
stb[at]uni-paderborn.de


Files:
FiS_Juli2012.pdf439 K