Combatting Online Extremism: Algorithmic Amplification of Counter-Speech.

恐怖分子和暴力极端分子历来利用社交媒体分享恐怖主义和暴力极端主义内容(TVEC)，这些内容在某些情况下可以通过算法放大(Alvari等人，2019年)。因此，建议利用算法来扩大针对目标受众的反演讲，作为打击暴力极端主义(CVE)的战略(Schmitt等人，2018年)。这个博士将探索围绕反言语算法放大的有效性和伦理。通过对研究问题的探讨，本研究旨在更好地了解反言语在劝阻个人使用极端叙事方面的有效性，找出反言语测量、监测和评估(MME)的挑战，并确定算法在向目标受众放大反言语方面的有效性和局限性。此外，这项研究将确定应支持反演讲的算法放大的关键伦理原则，并确定目前支持竞选活动的伦理原则。这项研究对于告知政策制定者、从业者和科技公司如何在道德上和有效地使用算法来放大反言论、劝阻个人不要发表极端的叙述具有重要意义。考虑到最近的监管努力，这一点尤其及时，《在线安全法案》就是证明。此外，这项研究将确定资源和研究可以在哪里解决现有的挑战。这将有助于打击网络极端主义，并缓解与反言论相关的潜在负面影响，包括强化极端主义观点和加剧两极分化。

Terrorists and violent extremists have historically used social media to share terrorist and violent extremist content (TVEC), which can be algorithmically amplified in certain contexts (Alvari et al., 2019). Therefore, it is suggested that algorithms should be utilised to amplifycounter-speech to target audiences, as a countering violent extremism (CVE) strategy (Schmitt et al., 2018). This PhD will explore the effectiveness and ethics surrounding the algorithmic amplification of counter-speech. By exploring the research questions, this research aims to gain a better understanding of the effectiveness of counter-speech at dissuading individuals from extreme narratives, identify counter-speech measurement, monitoring and evaluation (MME) challenges, and determine the effectiveness and limitations of algorithms in amplifying counter-speech to targetaudiences. Additionally, this research will identify the key ethical principles that should underpin the algorithmic amplification of counter-speech, and determine what ethical principles currently underpin campaigns. This research is important to inform policymakers, practitioners, and tech companies of how to ethically and effectively use algorithms to amplify counter-speech, to dissuade individuals from extreme narratives. This is particularly timely given recent regulatory efforts as evidenced by the Online Safety Bill. Furthermore, this research will identify where resourcesand research can address existing challenges. This will aid in combatting online extremism, and mitigate against potential negative impacts associated with counter-speech including; reinforcing extremist views, and increased polarisation.