Data Science in Digital Environments

Dies ist ein vergangenes Event

80 Personen haben teilgenommen

Bild des Veranstaltungsortes

Details

Data is created in vast amounts and at an incredible speed nowadays. With 500 million tweets, 65 billion WhatsApp messages, and more than 720,000 hours of new YouTube content being shared every single day, digital environments play a key role in this process. This leads to two fundamental questions for CorrelAid and data science for social good in general: (1) How can we analyze this huge amount of digital trace data? (2) How can society benefit from our analyses?

In order to tackle these questions, we've put together a diverse program of talks, starting with the keynote speech on the #MeTwo movement. Together with social activist Ali Can, who launched the hashtag last summer to draw attention to racial discrimination and racist attacks, data scientists Juan Orduz, Sandra Meneses, Paul Meiners, Konstantin Gavras, and Lisa Hehnke analyzed more than 200,000 tweets on #MeTwo. Using NLP techniques and dynamic network visualizations, they identified important topics and influential actors of the Twitter campaign. In their keynote, Ali and Juan will share the results of the project and discuss the societal benefits and limitations of such online movements with the audience.

Sociologist Vadim Voskresenskii will present a project he did together with his colleagues from the Weizenbaum Institute for the Networked Society. They analyzed transnational dynamics of the "120 decibels" campaign, launched by members of the Austrian Identitarian Movement in early 2018, as an attempt to latch onto an existing public agenda item, while subverting and undermining the goals of the original movement. These "hijacking" dynamics were marked by a narrowing of the scope of problem definition, in order to focus specifically on acts of violence against women by foreign men, and a reformulation of political demands aimed at European immigration policies. Vadim and his colleagues drew on user generated data from social media (Twitter, Facebook, and YouTube) in order to investigate the transnational scope of these coordinated efforts and its interaction with the MeToo movement via geolocation data and language clusters.

Mathematician Liubov Tupikina from CRI Paris will talk about the potential of network analysis for good. In her talk titled "Network analysis: How can we benefit from our travels?" she will first demonstrate how to analyze real-world heterogeneous transportation networks. She will then present Lecturers without borders (https://scied.network/), a project which Liubov and her colleagues developed together with researchers from Germany, France, and Switzerland.

With mobility data we can gather valuable insights on the movement pattern within a city. Computer scientist Alexandra Kapp from the Technologiestiftung Berlin will talk about her project on collecting and analyzing millions of bike sharing trips: How to efficiently gather, store, and analyse the data.

[Unfortunately, Fabio Votta's talk on personalisation algorithms and extremist content online had to be canceled. You can find his slides for a similar talk at https://xrw-and-algorithms.netlify.com and you can also reach him on Twitter (@favstats) to discuss the topic.]

Please note that some of the talks and the discussion on #MeTwo will be in German.

*Agenda*
6:30 p.m. - 7 p.m. Doors open & networking
7 p.m. - 7:15 p.m. Introduction to the event and #MeTwo by Lisa Hehnke, Konstantin Gavras, and Ali Can
7:15 - 7:45 p.m. Keynote and discussion on #MeTwo
7:45 p.m. - 8:30 p.m. Talks by Vadim Voskresenskii, Liubov Tupikina, and Alexandra Kapp
8:30 p.m. - 8:45 p.m. Lightning talks by Malte Höpfner & Ellen König (TolaData), Juan Orduz (TD Reply), and Frie Preu (codecentric)
8:45 p.m. - 9 p.m. Summary and final statements
from 9 p.m. - Open end with drinks and snacks