Online Library

Extreme, anti-establishment actors are being characterized increasingly as ‘dangerous individuals’ by the social media platforms that once aided in making them into ‘Internet celebrities’. These individuals (and sometimes groups) are being ‘deplatformed’ by the leading social media companies such as Facebook, Instagram, Twitter and YouTube for such offences as ‘organised hate’. Deplatforming has prompted debate about ‘liberal big tech’ silencing free speech and taking on the role of editors, but also about the questions of whether it is effective and for whom. The research reported here follows certain of these Internet celebrities to Telegram as well as to a larger alternative social media ecology. It enquires empirically into some of the arguments made concerning whether deplatforming ‘works’ and how the deplatformed use Telegram. It discusses the effects of deplatforming for extreme Internet celebrities, alternative and mainstream social media platforms and the Internet at large. It also touches upon how social media companies’ deplatforming is affecting critical social media research, both into the substance of extreme speech as well as its audiences on mainstream as well as alternative platforms.

VOX-Pol Blog Post.

Mass transit systems are relied on a daily basis to transport millions of passengers and bring billions of dollars’ worth of economic goods to market. While some forms of mass transit rely on a fuel, electrified railway systems are dependent on the electric grid. The electric grid is becoming more vulnerable to disruptions, due to extreme weather, changing supply and demand patterns, and cyber-terrorism. An interruption to the energy supply of a railway infrastructure can have cascading effects on the economy and social livelihood. Resilient interconnected microgrids are proposed to maintain reliable operation of electri_ed railway infrastructures. An engineering design framework, and supporting methods and techniques, is proposed for an electrified railway infrastructure to be upgraded from its existing form, to one with resilient interconnected microgrids. The sizing of the interconnected microgrids is performed using an iterative sizing analysis, considering multiple resiliency key performance indicators to inform the designer of the trade-o_s in sizing options. Hierarchical control is proposed to monitor and control the interconnected microgrids. A multi-objective problem cast in the tertiary level of control is proposed to be solved using game theory. The proposed designs are modelled and simulated in Simulink. Four case studies of railway infrastructures in Canada and the United Kingdom are used to demonstrate the effectiveness of the proposed designs. While results for each case study vary, resilient interconnected microgrids for railway infrastructures demonstrates a reduced dependence on the electric grid. The examples here are all scalable and can perform within the framework of any available energy system. The results are both extremely impressive and promising towards a more resilient and stable energy future for our railway and other critical infrastructures.

Polarization is implicated in the erosion of democracy and the progression to violence, which makes the polarization properties of large algorithmic content selection systems (recommender systems) a matter of concern for peace and security. While algorithm-driven social media do not seem to be a primary driver of polarization at the country level, they could be a useful intervention point in polarized societies. This paper examines algorithmic depolarization interventions aimed at transforming conflict: not suppressing or eliminating conflict, but making it more constructive. Algorithmic intervention is considered at three stages: what content is available (moderation), how content is selected and personalized (ranking), and content presentation and controls (user interface). Empirical studies of online conflict suggest that not only could the exposure-diversity intervention proposed as an antidote to ‘filter bubbles’ be improved: under some conditions, it can even worsen polarization. Using civility metrics in conjunction with diversity in content selection may be more effective. However, diversity-based interventions have not been tested at scale, and may not work in the diverse and dynamic contexts of real platforms. Instead, intervening in platform polarization dynamics will likely require continuous monitoring of polarization metrics, such as the widely used ‘feeling thermometer’. These metrics can be used to evaluate product features, and can potentially be engineered as algorithmic objectives. While using any metric as an optimization target may have harmful consequences, to prevent optimization processes from creating conflict as a side effect it may prove necessary to include polarization measures in the objective function of recommender algorithms.

VOX-Pol Blog post.

In recent years, the increasing propagation of hate speech on social media and the urgent need for effective counter-measures have drawn significant investment from governments, companies, and empirical research. Despite a large number of emerging scientific studies to address the problem, a major limitation of existing work is the lack of comparative evaluations, which makes it difficult to assess the contribution of individual works. This paper introduces a new method based on a deep neural network combining convolutional and gated recurrent networks. We conduct an extensive evaluation of the method against several baselines and state of the art on the largest collection of publicly available Twitter datasets to date, and show that compared to previously reported results on these datasets, our proposed method is able to capture both word sequence and order information in short texts, and it sets new benchmark by outperforming on 6 out of 7 datasets by between 1 and 13% in F1. We also extend the existing dataset collection on this task by creating a new dataset covering different topics.