Im Rahmen des Ed-Tech Research Forums hielt Alexander Christ einen Workshop zum Thema „Big Data Methods and Text-Mining for Systematic Reviews“. Am Beispiel unserer Forschungssynthesen an der Schnittstelle von Digitalisierung und kultureller Bildung im DiKuBi-Projekt veranschaulichte er die Relevanz von Big Data Methoden für das Feld der kulturellen Bildung und für die Bildungsforschung im weiteren Sinne zeigen.
Systematic reviews are the method of choice to synthesise research evidence. To identify topics with many original publications that are in need of a synthesis (so-called “hot spots”) it is necessary to cope with the three V’s of big data (Volume, Velocity and Variety), especially in loosely defined or fragmented disciplines. For this purpose, text mining methods may substantially facilitate the analysis of large literature corpora. Thus, we applied text mining and predictive modelling to digitalisation in aesthetic, arts and cultural education (D-ACE), and particularly to quantitative research in D-ACE (QRD-ACE) as a prototypical loosely defined, fragmented discipline. Applying a broad search query of terms indicative of QRD-ACE to Scopus, we identified a corpus of N = 55,553 publications for 2013–2017. As the result of an iterative approach of text mining, manual screening and predictive modelling, we identified n = 8,304 potentially relevant publications of which n = 1,666 were included after manual screening. Analysing the distribution of the included
publications revealed video games as a major hot spot of QRD-ACE (n = 836). Topic modelling revealed self-representation in social networks as a second hot spot related to four out of k = 8 identified topics. Text mining and predictive modelling shortened the screening time to less than 20% compared to screening the whole corpus. We discuss implications for harnessing text mining in future research syntheses and for future original publications regarding quantitative research on D-ACE.