%0 Conference Proceedings
%T Using parallel distributed processing to reduce the computational time of digital media similarity measures
%+ George Mason University [Fairfax]
%A Lim, Myeong
%A Jones, James
%Z Part 2: Approximate Matching Techniques
%< avec comité de lecture
%( IFIP Advances in Information and Communication Technology
%B 17th IFIP International Conference on Digital Forensics (DigitalForensics)
%C Virtual, China
%Y Gilbert Peterson
%Y Sujeet Shenoi
%I Springer International Publishing
%3 Advances in Digital Forensics XVII
%V AICT-612
%P 65-87
%8 2021-02-01
%D 2021
%R 10.1007/978-3-030-88381-2_4
%K Drive similarity
%K sector hashes
%K Jaccard index
%K parallel computation
%Z Computer Science [cs]Conference papers
%X Digital forensic practitioners are constantly challenged to find the best allocation of their limited resources. While automation will continue to partially mitigate this problem, the preliminary question about which media should be prioritized for examination is largely unsolved. Previous research has developed methods for assessing digital media similarity that may aid in prioritization decisions. Similarity measures may also be used to establish links between media and, by extension, the individuals or organizations associated with the media. However, similarity measures have high computational costs that delay the identification of digital media warranting immediate attention and render link establishment across large collections of data impractical.This chapter presents and validates a method for parallelizing the computations of digital media similarity measures to reduce the time requirements. The proposed method partitions digital media and distributes the computations across multiple processors. It then combines the results as an overall similarity measure that preserves the accuracy of the original method executed on a single processor. Experiments on a limited dataset demonstrate reductions of up to 51% in processing time. The reductions vary based on the number of partitions chosen and specific digital media being examined, suggesting the need for additional testing and optimization strategies.
%G English
%2 https://inria.hal.science/hal-03764373/document
%2 https://inria.hal.science/hal-03764373/file/522103_1_En_4_Reference.pdf
%L hal-03764373
%U https://inria.hal.science/hal-03764373
%~ IFIP-LNCS
%~ IFIP
%~ IFIP-AICT
%~ IFIP-TC
%~ IFIP-WG
%~ IFIP-TC11
%~ IFIP-DF
%~ IFIP-WG11-9
%~ IFIP-AICT-612