Vortrag: Philipp Heinrich (11.12.2024)
Im Rahmen des Oberseminars Computerlinguistik findet am 11.12.2024 ein Vortrag statt, zu dem wir herzlich einladen möchten.
Vortragender:
Philipp Heinrich (CCL FAU)
Zeit:
Mittwoch, 11.12.2024, 16:15-17:45 Uhr
Ort:
Bismarckstr. 12, R.0.320 (05204.00.320)
Thema:
“Analysing Discourse with Discoursemes”
Abstract:
For the last decades, corpora have played a crucial role in countering epistemic relativism in discourse analysis. Corpus-assisted Discourse Studies (CADS) try to bring together empirical methods such as keyword and collocation analysis and the qualitative discussion and critique of societal issues (which are often rooted in Critical Discourse Analysis). Successful CADS studies identify linguistic patterns (such as words or n-grams) that are indicative of discursive patterns (such as claims or attitudes). In recent publications (Heinrich et al. 2018, Heinrich & Evert 2024), we developed the concept “discourseme” as (minimal) unit of lexical meaning in the context of a given discourse. Discoursemes combine an intensional meaning description at the qualitative level with an operationalisation in terms of lexical items that can be queried in corpora and used for further quantitative analyses. In this talk, I give an overview of the functionality of cwb-cads (https://github.com/ausgerechnet/cwb-cads), a dedicated REST API for analysing discourses with discoursemes, with a focus on visualisations in terms of semantic maps (Evert & Heinrich 2019) and discourseme networks. I will also have a look at the role that corpus queries which go beyond simple unigrams or multi-word units can play in discourse analysis.
References:
– Evert, Stefan and Heinrich, Philipp (2019). Introducing MMDA: An interactive toolkit for CDA. Presentation at the 7th Göttingen-Hildesheim-Workshop on Computational Linguistics and Digital Humanities: Digital Methods in Political Science. Göttingen, Germany.
– Heinrich, Philipp; Adrian, Christoph; Kalashnikova, Olena; Schäfer, Fabian; and Evert, Stefan (2018). A Transnational Analysis of News and Tweets about Nuclear Phase-Out in the Aftermath of the Fukushima Incident. In Proceedings of the LREC 2018 “Workshop on Computational Impact Detection from Text Data”, page 8–16, Paris. ELRA.
– Heinrich, Philipp; Evert, Stephanie (2024). Operationalising the Hermeneutic Grouping Process in Corpus-assisted Discourse Studies. In Proceedings of the 4th Workshop on Computational Linguistics for the Political and Social Sciences: Long and short papers, pages 33–44, Vienna, Austria.