Dr. Jessie Labov
8 November 2019, 12:00 am
Instytut Informacji Naukowej i Bibliotekoznawstwa UWr 
pl. Uniwersytecki 9/13
Room 104
 

As the term ‘digital humanities’ is used to describe more and more varied activities, from data-driven research in humanities disciplines to public-facing interactive exhibits at cultural heritage institutions, it is increasingly difficult to track any one direction in which it is developing, or to point to specific disciplinary trends. At best, we might say that from the mid-2000s until mid-2010s, most digital history projects involved relating entities via maps or graphs, which extended the scope of small-scale studies onto larger historical stage. However, the rapid evolution over the last few years of machine-learning enhanced analysis in linguistics and literature is beginning to have an impact in other areas. Now historians are becoming more interested and educated in textual analysis; while social scientists realize that some of their most basic methodologies which rely on human coding could be transformed via text processing and mining. The question which repeatedly comes up, however, with every new experiment with text mining on the frontiers of new disciplines, is whether these same tools of computational linguistics can be employed in the new context. This talk will draw from the examples of two projects, Reading Kultura from a Distance and Planet Telex, which are poised between the two possibilities – metadata extraction and full-scale text mining – to explain what happens when this frontier is crossed, and what the implications are for disciplines far beyond linguistics and philology.