Page 46 - Understandinging Forensic Technology Landscape

Supervised learning: Instruction of a ML model that Write blocker: A hardware write blocker (HWB) is a
involves mapping an input to a result based on example hardware device that attaches to a computer system
pairings in a training data set. Supervised learning with the primary purpose of intercepting and preventing
generally requires multiple iterations of training where (or ‘blocking’) any modifying command operation from
a ML model gets feedback from a human reviewing ever reaching the storage device. Physically, the device is
results. connected between the computer and a storage device.
Some of its functions include monitoring and filtering
Unstructured data: Unstructured data has internal any activity that is transmitted or received between its
structure but is not structured via pre-defined data interface connections to the computer and the storage
models or schema. It may be textual or non-textual, and device.
64
human- or machine-generated.
63
Unsupervised learning: Instruction of a ML model that
involves making inferences from data without a training
set of input-response pairs. The most common form of
unsupervised learning is cluster analysis, which finds
hidden patterns/groupings in data using measures of
similarity.

63 See footnote 59
64 Hardware Write Blocker Device (HWB) Specification, National Institute of Standards and Technology, nist.gov/system/files/documents/2017/05/09/hwb-v2-post-19-may-04.pdf,
accessed March 18, 2020

Understanding the forensic technology landscape | 42