Research Area:  Digital Forensics

Abstract:

Digital multimedia forensics (DMF) plays an outstanding role in digital forensics (DF) area as it encompasses solutions for promoting image, audio, and video evidence to be used in court. As every science, one straightforward way of evolving is conducting formal experiments and providing reliable and auditable evidence. Therefore, in this article, we systematically map and analyze how experiments in DMF have been conducted and whether data-based evidence has been provided. This article presents a subset of results on DMF from a wider mapping study on DF experimentation. We then analyzed 49 experiments resulting from electronic and manual searches in widely known electronic databases and conferences/workshops on DF. We used them to answer six research questions. We demonstrated that DMF needs to improve the way experiments are reported, especially how data are shared to allow reproducibility and, consequently, evolution of the research topic. We show that experimental elements, for example, hypothesis, variables, design, instrumentation, validity evaluation, setup, training, data sets and benchmarks, statistical techniques (descriptive, hypothesis, and effect size test), and limitations, should be better reported. As a general result, we understand DMF formal experimentation needs to evolve toward providing useful, reliable, auditable, and reproducible experimental evidence as both academia and practitioners might straightforward benefit from such evolution.

Keywords:  

Author(s) Name:  Edson Oliveira Jr, Avelino F. Zorzo, Charles V. Neu

Publisher name:  Wiley

DOI:  https://doi.org/10.1002/wfs2.1405

Volume Information:  Volume3, Issue4

Paper Link:   https://wires.onlinelibrary.wiley.com/doi/abs/10.1002/wfs2.1405