Do we live in online bubbles ? EPFL Research

[…] Scientists from EPFL’s Data Science Lab (dlab) in the School of Computer and Communication Sciences, with colleagues from MIT and the Mozilla Foundation have analyzed the browsing and reading habits of tens of thousands of people using a plugin used with the Firefox browser, for the first time gaining a glimpse at polarization from a content consumption perspective. 

Echo chambers are not a new thing but the real replication of these memes happens online. Until now, a lot of polarization research has been around content production because this is what we can easily measure. This population of Firefox users consented to share their data – it’s like some people donate blood for the common good, here people donated their data for the common good,” said Head of the Data Science Lab, Assistant Professor Robert West, and the study’s lead author.

Whilst there have been previous, smaller studies that measured engagement in different ways, uniquely this new study was conducted in vivo, with users going about their normal daily lives, meaning that researchers were able to follow people in their natural habitat. With access to browsing history, unlike earlier studies, this research measured the time that users spent on particular websites, and reading particular articles, rather than whether a user had visited a site or not. 

This additional data provided new evidence of a greater extent of polarization than observed in prior literature, showing that people engaged much more deeply with articles matching their political persuasion, spending more time on news sources matching their partisan beliefs than other information sources. […]

Source and full article : EPFL

Forschungsübersicht zu Künstlicher Intelligenz

Als Basistechnologie, die in allen Bereichen des täglichen Lebens Einzug findet, hat KI eine herausragende Bedeutung für den Forschungsstandort und die zukünftige Wertschöpfung in der Schweiz. Die geplante Plattform – die Swiss Artificial Intelligence Research Overview Platform (SAIROP) – soll eine Übersicht über die KI-Kompetenzen in der zerstreuten Forschungslandschaft der Schweiz darstellen. SAIROP wird die Sichtbarkeit von Schweizer KI-Forschungspartnern verbessern und damit für lokale und internationale Unternehmen und Forschungseinrichtungen einen neuen Zugang für die Initiierung zukünftiger Innovationsprojekte bieten.

Die involvierten Partner aus Forschungsförderung, Wissenschaft und Industrie tragen Informationen zu aktuellen Forschungsprojekten aus bestehenden Datenbanken zusammen, bereiten diese auf und machen sie mittels der neuen Plattform besser zugänglich – was in dem geplanten Detaillierungsgrad im europäischen Raum einmalig sein dürfte. Hier kommt der Connectome Knowledge Graph zum Einsatz: die Daten unterschiedlicher Quellen werden im Hintergrund verlinkt und ermöglichen so über SAIROP als zentraler Zugangspunkt die Kompetenzen Schweizer Forschungsgruppen anhand von aktuellen Projekten aufzeigen. SAIROP soll im Herbst 2021 lanciert werden und öffentlich zugänglich sein.

Zur vollständigen Pressemitteilung der SATW zu SAIROP.

Source : SWITCH

The Evolving Cyber Threat Landscape during the Coronavirus Crisis – CSS Cyberdefense Report

In light of the societal changes wrought by the coronavirus pandemic, this report aims to examine the impact this crisis had on the general cybersecurity landscape, notably in terms of the attack surface and exploitation opportunities; to investigate the changing and recurring patterns of adversarial behaviors; and to illustrate and provide an overview of how threats actors have leveraged said epidemic in Q1/Q2 2020.

Accordingly, this report highlights that the coronavirus pandemic has generated a set of remarkable and psycho-societal, technical, and logistical economic circumstances upon which malicious actors have capitalized. These factors include:

1) an expanded socio-technical attack surface due to the greater use and dependency on services and applications for telework provided through digital infrastructure in general and cloud infrastructure in particular;

2) a psycho-informational environment characterized by anxiety, uncertainty, and high demand for information;

3) a nexus of economic and trade uncertainty/ disruption, emergency procurement processes, compounded by the wide availability of nefarious cyber tools.

23122020_CyberThreatLandscapeCoronaCrisis

Source : CSS ETHZ