Data

Historische Medien spiegeln die Werte- und Wissenshorizonte vergangener Zeiten wider. Für ihre Erforschung werden auch zunehmend Methoden des maschinellen Lernens genutzt. In den letzten Jahren sind neben Volltextsuchen weitere Methoden hinzugekommen: Topic modeling, Bildähnlichkeitsanalysen, Text Reuse Detection, Spracherkennung, oder Word Embeddings eröffnen jeweils neue Zugänge zu großen Textsammlungen.

In the impresso project, a team of computational linguists, historians and designers strive to enable data-driven analyses of historical media across time, institutional silos, media types and languages. To this end we compile a corpus of historical newspaper and radio collections with the help of Western European partners, enrich it using text mining techniques and develop user interfaces for their exploration and computational analysis.

Many network analysis projects rely on somewhat ready-made sources for data; for example, email logs, questionnaires, church registers, letter exchanges and trade relations make it relatively easy to identify who is connected to whom and how. It is, however, considerably more difficult to extract quantifiable data from text. Some issues to consider here are: how can we bridge the gap between the depth of hermeneutics and data analysis? How can we systematize text interpretation?

Le 8 mai dernier, Emmanuel Macron se tenait seul au bout d’une avenue vidée, de peur d’une possible casserolade. Cette image marquante nous rappelle à considérer les commémorations comme des enjeux politiques de premier ordre, renvoyant toujours à un état des relations entre État, pouvoir et citoyens, y compris dans ce que ce lien peut avoir de plus conflictuel. Il en ira des mêmes enjeux pour la commémoration, ce 11 novembre, des 105 ans de l’armistice de 1918.

This presentation aimed at reflecting on the history of vernacular digital cultures and their historical approach through web archives. It relied on the experience of the Hivi project (A history of Online Virality) and notably the historical study of memes we conducted. This led to envisioning some desirable futures in approaching vernacular digital cultures (issues of access, inclusiveness, role of visual and sound studies, readability of web archives ,etc.)

This lightening talk is part of the panel "Teaching history of technology: experiments, innovative methods, and approaches". It aims to discuss the experiment Prof. Schafer conducted over 14 weeks with her master students in creating a Living Book intertwining sources and research articles on the topic “Women, Gender and ICTs”.

Luxembourg is a de facto multicultural country, with 179 different nationalities represented. It is, however, complex to identify who among the latter is perceived as an immigrant by public opinion. In the same vein, immigration stories rarely make the headlines of some of the most prominent outlets of Luxembourg’s mainstream media. This study covers the print content of some of Luxembourg’s dominant media outlets in the search for the representation of (im)migrants and refugees.

The military offensive launched by the Russian Federation against Ukraine in the early hours of 24 February 2022 and the ensuing war on Ukrainian territory have resulted in immense human suffering, a humanitarian tragedy and incalculable material damage.

Workshop for the EUROPAST project Lund workshop.

Die Schueberfouer ist der größte Jahrmarkt in Luxemburg, der alljährlich gegen Ende des Sommers auf dem Glacis Feld in der Hauptstadt stattfindet. In Luxemburg haben zahlreiche Menschen persönliche Erinnerungen an Besuche und Erlebnisse auf diesem Jahrmarkt, und es wird vielfach über seine historische Bedeutung geschrieben, dessen Ursprünge bis ins Jahr 1340 zurückreichen.