In the April edition of DRS seminar, Pihla Toivanen introduced her Master’s thesis “Computational Frame Analysis of Populist Counter Media”. Her research is a part of the project “Information chaos and trust in traditional journalism” project carried out at Tampere University’s COMET Research Center in collaboration with Aalto University.
Pihla uses machine learning to study how Finnish populist media builds trust by referring to external information sources. In particular, populist media actively use news coming from traditional media to produce their own content (see table below). Mainstream media in Finland are trusted, and reusing their stories may help populist media to build credibility. However, populist media do not merely reproduce information from the traditional media, but actively reframe it. Pihla developed a supervised machine-learning algorithm that detects how populist media reframe stories from traditional media. The algorithm and the preliminary results of its application to unlabeled data were presented at the seminar, as well as the researcher’s future plans to develop the classification technique. After the presentation, seminar participants discussed the use of machine learning in media studies and new methodologies of media research.