ABSTRACT

Social media platforms and news organisations alike are struggling with identifying and combating visual mis/disinformation presented to their audiences. Such processes are complicated due to the enormous number of media items being produced, how quickly media items spread, and the often-subtle or sometimes invisible-to-the-naked-eye nature of deceptive edits. Despite knowing little about the provenance and veracity of the visual content they encounter, journalists have to quickly determine whether to re-publish or amplify this content, with few tools and little time available to assist them in such an evaluation. With the goal of equipping journalists with the mechanisms, skills, and knowledge to be effective gatekeepers and stewards of the public trust, this study reviews current journalistic image verification practices, examines a number of existing and emerging image verification technologies that could be deployed or adapted to aid in this endeavour, and identifies the strengths and limitations of the most promising extant technical approaches. While oriented towards practical and achievable steps in combating visual mis/disinformation, the study also contributes to discussions on fact-checking, source-checking, verification, debunking and journalism training and education.

摘要

社交媒体平台和新闻机构都在努力识别和打击呈现给观众的视觉错误/虚假信息。由于制作的媒体项目数量巨大，媒体项目传播的速度有多快，以及欺骗性编辑通常微妙或有时肉眼不可见，因此此类过程很复杂。尽管对他们遇到的视觉内容的出处和真实性知之甚少，但记者们必须迅速确定是否重新发布或放大这些内容，而帮助他们进行此类评估的工具和时间也很少。为了让记者掌握机制、技能和知识，成为公众信任的有效守门人和管家，本研究回顾了当前的新闻形象验证实践，研究了许多现有和新兴的图像验证技术，这些技术可以部署或调整以帮助这项工作，并确定最有前途的现有技术方法的优势和局限性。该研究旨在打击视觉错误/虚假信息的实际和可实现的步骤，同时也有助于讨论事实核查、来源核查、核实、揭穿以及新闻培训和教育。