Harnessing the Power of Text Mining for Humanities Research.

Textual data have always been essential to research and education within the humanities. The standard humanities approach to textual data is a combination of qualitative methods and human synthesis, that is, humanist researchers apply close readings and theoretically motivated arguments to a given set of textual data with the purpose of understanding the meaning of those data. The increasing availability of digital texts and computing power is beginning to have an impact on this standard approach. Humanities researchers are starting to ask new types of questions to old textual problems and to propose novel solutions based on data intensive scientific discovery. These solutions use faster and more efficient methods to collect, analyze, and visualize textual data.

This talk will consist of two main parts: 1) a short introduction to data intensive discovery using unstructured data sets (i.e., text mining) from the perspective of digital humanities. Text mining covers several related methods for processing textual data, all of which need to convert ‘text to numbers’. In the humanities there are essentially two interrelated arguments for applying text analytics, the qualitative argument (can we discover something new?) and the quantitative argument (can we discover something faster?); 2) a detailed presentation of several projects using text mining at Aarhus University. 

Kristoffer Laigaard Nielbo, Assistant Professor, Aarhus University, Study of Religion.