Study of duplicate tuple management

Data management efficiency is a crucial factor of current processing. Information systems now require huge data amount to be handled, evaluated to cover robustness and decision making. Input data can originate from various sources and can consist of multiple duplicate tuples, which must be filtered out. This paper aims to discuss methods for such tuple identification, either in the input stream, index access path, as well as during the result set composition. The core part is ensured by our new proposed technique of the multi-indexing layer. The contribution of the paper takes emphasis to the Hash Epsilon approach in the index access, whereas the different precision can be present. Thanks to applying proposed methods, storage demands can be lowered ensuring data consistency and references.