With the explosion of data stored by individuals and organizations, deduplication and compression techniques are becoming ever more popular. These techniques are commonly referred to as data reduction and are studied (among other techniques) in the context of storage systems. As part of IBM's initiative to support data reduction in its storage offerings, our group works on various aspects of this problem, both from the research angle as well as in supporting our products.
We developed an in-depth understanding into the problem of estimation of data reduction, both for compression and for deduplication.
We work on full-object deduplication, particularly for the cloud. In addition, we perform studies on security issues that arise from performing over-the-wire deduplication.