Laying foundations for effective machine learning in law enforcement. Majura – A labelling schema for child exploitation materials

Dalins, Janis; Tyshetskiy, Yuriy; Wilson, Campbell; Carman, Mark James; Boudry, Douglas

doi:10.1016/j.diin.2018.05.004

Cited by 21 publications

(8 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Facebook and major content providing and hosting services have taken steps to strengthen their algorithms; for example, in August 2019, Facebook open-sourced their algorithms for video and image detection. 119 There have been no significant developments or improvements in hashing or AI detection since the Christchurch Massacre. 120 The current AI technology and human-based frameworks are unable to remove all offending content after notification, even over several weeks.…”

Section: Impossible Burdenmentioning

confidence: 99%

The Sharing of Abhorrent Violent Materials Act: The realities and implications of Australia’s new laws regulating social media companies

Valcic

2021

Bond Law Review

View full text Add to dashboard Cite

On 15 March 2019, a Facebook Live video was broadcast from Christchurch, New Zealand, documenting a terror attack which resulted in the death of fifty-one people. This attack highlighted a weakness in social media protections and a gap in legislation globally. In response, the Criminal Code Amendment (Sharing of Abhorrent Violent Material) Act 2019 (Cth) seeks to make internet service and social media providers accountable for the removal of abhorrent and violent content. This legislation sent waves through the international community, attracting criticisms for its fast adoption, perceived unrealistic obligations and harsh penalties, as well as its broad extraterritorial reach. This article will explore these criticisms. It asks, how is the Act exercising extraterritorial jurisdiction? Is there an unrealistic burden created? And if there is a breach, who will be charged? The article concludes that if these challenges are not adequately addressed, the Act will not only fail to achieve its goal of reducing the accessibility of abhorrent violent material, but will also pose a serious threat to the protection of human rights.

show abstract

Section: Impossible Burdenmentioning

confidence: 99%

The Sharing of Abhorrent Violent Materials Act: The realities and implications of Australia’s new laws regulating social media companies

Valcic

2021

Bond Law Review

View full text Add to dashboard Cite

show abstract

“…Studies Online detection [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [9], [39], [40], [41], [42], [43], [8], [44], [45], [46], [47], [48], [49] Offline detection [50], [51], [52], [53], [54], [7], [55], [56] Safety [57], [13], [58], [59], [60], [61],…”

Section: Categorymentioning

confidence: 99%

“…These studies built ML models that received images, videos, [8] written text [9], or a combination of them [28] as input, to automatically classify them as abusive or not abusive towards women or children. ML used in this context made it possible to quickly separate large amounts of abusive Internet content from its non-abusive counterpart, a task that is often time-consuming [42] or emotionally-taxing [46] when done manually. In order to detect and eliminate VAW and VAC, it is necessary to address explicit cases of violence, but also the risk factors that lead to their occurrence.…”

Section: ) Online Detectionmentioning

confidence: 99%

A Systematic Review of Computer Science Solutions for Addressing Violence Against Women and Children

et al. 2021

View full text Add to dashboard Cite

Violence against women and children are public health issues of pandemic proportions. It is estimated that one in every three women worldwide have experienced physical, emotional or sexual violence. Similarly, each year one of two children are victims of some form of violence, including domestic aggression and bullying. Due to the widespread use of the Internet and social media, women and children are also now vulnerable to other types of violence such as cyber-bullying and online sexual or emotional harassment. To help alleviate these social problem, the use of computer sciences and related technologies has been leveraged in recent years. The Internet of Things, artificial intelligence, ubiquitous and mobile computing, pattern recognition, cloud computing, and similar technologies have been used to formulate solutions to detect and prevent violent acts. In this paper, a systematic review of some of the efforts that can help address the problem of violence against women and children is presented. This paper describes the current state-of-the-art of these contributions and identifies trends, architectures, technologies, and current open challenges. The survey was developed using a literature review of academic documents published from 2010 to 2020. The contributions were categorized in four application domains: online detection, offline detection, safety and education. Additionally, these contributions were further categorized based on the computer science approaches and technologies used: artificial intelligence, Internet of Things and digital serious games.

show abstract

“…These matters may remain subject to further judicial proceedings, therefore no further details of their provenance will be provided. For tests benefitting from separate corpora, we utilised about 322,490 images of lawful pornography first used by [6], plus a partial (approx. 596,000 images) download of the Google Open Images Dataset [7].…”

Section: Test Corpusmentioning

confidence: 99%

PDQ & TMK + PDQF -- A Test Drive of Facebook's Perceptual Hashing Algorithms

Dalins¹,

Wilson²,

Boudry³

2019

Preprint

Self Cite

View full text Add to dashboard Cite

Efficient and reliable automated detection of modified image and multimedia files has long been a challenge for law enforcement, compounded by the harm caused by repeated exposure to psychologically harmful materials. In August 2019 Facebook open-sourced their PDQ and TMK + PDQF algorithms for image and video similarity measurement, respectively. In this report, we review the algorithms' performance on detecting commonly encountered transformations on real-world case data, sourced from contemporary investigations. We also provide a reference implementation to demonstrate the potential application and integration of such algorithms within existing law enforcement systems.

show abstract

Laying foundations for effective machine learning in law enforcement. Majura – A labelling schema for child exploitation materials

Cited by 21 publications

References 16 publications

The Sharing of Abhorrent Violent Materials Act: The realities and implications of Australia’s new laws regulating social media companies

The Sharing of Abhorrent Violent Materials Act: The realities and implications of Australia’s new laws regulating social media companies

A Systematic Review of Computer Science Solutions for Addressing Violence Against Women and Children

PDQ & TMK + PDQF -- A Test Drive of Facebook's Perceptual Hashing Algorithms

Contact Info

Product

Resources

About