The Evidentiary Value of Big Data Analysis

Abstract

Big data is transforming the way that governments provide security to, and justice for, their citizens. It also, however, has the potential to increase surveillance and government power. Geospecific information – from licence plate recognition and mobile phone data, biometric matches of DNA, facial recognition, financial transactions, and internet search history – is increasingly allowing government agencies to search and cross-reference. This increased reliance on big data searches raises the question: what is the probative value of the information that results? A distinguishing feature of the scientific method is the development of a hypothesis that is then tested against data that either support or refute the hypothesis. A conventional criminal investigation can be seen to parallel that method – after a suspect is identified, evidence is gathered to either build a case against, or rule out, that suspect. The analysis of big data, by contrast, can deviate from a path that parallels the scientific method, at times being more akin to trawling through data to then generate a hypothesis. In this paper we investigate conditions in which this leads to problematic outcomes, such as more data leading to high rates of false positives. We then sketch a big data analysis legal/policy framework that can circumvent these problems.



Author Information
Marco Pollanen, Trent University, Canada
Bruce Cater, Trent University, Canada

Paper Information
Conference: ACPEL2016
Stream: Law 12. Criminal Justice Policy and Law

This paper is part of the ACPEL2016 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window

Posted by amp21