Abstract

Data visibility is essential to providing effective data security. Identifying key data assets is a mandatory first step to providing protection, compliance, governance, and privacy.  IBM Security Discover and Classify enables customers to discover and classify data at scale.

IBM commissioned Tolly to benchmark its Security Discover and Classify, supervised-AI, data  discovery and classification solution. The evaluation included accuracy and throughput benchmarks for both structured (database) and unstructured (flat file) data. Additionally, tests were run on image data  and a Microsoft Exchange email environment.

Testing demonstrated that IBM Security Discover and Classify could deliver 98.6% accuracy in the test  of structured data and 100% accuracy in the test of unstructured data.  For throughput tests, appropriate metrics were developed to report the results in terms of data throughput and object throughput (e.g. images per hour.)