Big Data in the Chamber: Corpus-Assisted Studies of Parliamentary Discourse Across Time and Space

Instructors: Anna Kryvenko
Duration: both weeks

Abstract

Parliaments are pivotal institutions in democracies, shaping policies that impact citizens by deliberating critical societal issues. The debates are commonly recorded as open-access digital proceedings enriched with metadata. These records are valuable for researchers exploring political, societal, historical, cultural or communicational dynamics in fields such as linguistics, discourse analysis, political science, history, sociology, gender studies as well as for various teaching contexts.
This workshop takes advantage of the interoperability and comparability of the ParlaMint corpora containing parliamentary proceedings from 26 national and 3 regional parliaments across Europe at least between 2015 and 2022, although several ParlaMint corpora include data spanning a much longer period. Available in the original languages and machine-translated to English, the corpora also feature metadata on speakers, parties and speeches, including names, gender, age, roles, party affiliation, power positions, political leanings, speech dates, topics and sentiment. This hands-on project-oriented tutorial will provide skills and methodological training to explore ParlaMint version 5.0, which can be obtained by downloading the files or by accessing the preloaded data via online platforms – primarily noSketch Engine and TEITOK. All data and tools are open access and can be used free of charge.
Designed for researchers in Social Sciences and Humanities with interest in parliamentary discourse but no or little familiarity with corpus linguistic tools, this workshop will train participants to leverage extensive content, annotations and metadata via user friendly concordancers, facilitating research on individual national parliaments, enabling transnational comparisons, and fostering cross-disciplinary collaboration. Participants with also discover CLARIN – Common Language Resources and Technology Infrastructure.

Learning outcomes

Participants will discover CLARIN’s language resources and gain skills for querying ParlaMint corpora, including simple and advanced CQL queries. They will learn to analyse extracted results alongside speaker and text metadata, including sentiment and topic analysis under the ParlaCAP project. Additionally, they will learn to frame and interpret the results for various Digital Humanities and Social Sciences research questions, such as comparing speeches and attitudes of ruling and opposition parties or exploring parliamentary protocol variations. The techniques taught are transferable to other corpora and concordancers. The showcases will use ParlaMint-GB containing proceedings from the UK Parliament, and the machine-translated version of all other ParlaMint corpora into English (Parlamint_xx_en) to ensure all participants can follow queries and discuss results. However, for the hands-on activities and project work, the participants will be encouraged to use the ParlaMint corpora in other languages.

Datasets

ParlaMint version 5.0 can be downloaded from the CLARIN.SI repository as a TEI-encoded corpus, a linguistically marked-up corpus in several formats and a version machine-translated into English, all with added metadata files. For on-line analysis, different versions of the ParlaMint corpora are preloaded to various tooks, including noSketch Engine (up to v. 5.0) and TEITOK (v. 4.1).

Teaching materials

Digital textbook

Pahor de Maiti Tekavčič, K., & Kryvenko, A. (2026). From the dispatch box: Unlocking the potential of ParlaMint through noSketch Engine and TEITOK (Ed. 1.0) [E book]. Inštitut za novejšo zgodovino / Institute of Contemporary History.

Technical requirements

Organisers should provide an LCD projector with a HDMI output, WIFI and some power extension cords for the participants.
Participants should bring their own laptops. Participants are also advised to create a free personal account at https://www.clarin.si/skelog prior to the beginning of the workshop.

Brief course timeline

Week 1 (15 hours + an open teaser session)

Day 1

Introduction to CLARIN’s language data tools and resources
Introduction to the ParlaMint corpora and related projects
“Adopting” a ParlaMint corpus (corpora) and planning project work Daily reflection

Day 2

Studying parliamentary discourse via corpora. How verbatim are transcripts?
Assessing the ParlaMint datasets via user-friendly tools
Daily reflection

Day 3

ParlaMint provenance information, structural markup, linguistic annotation and metadata
Delimiting searches and contextualising results
Daily reflection

Day 4

An overview of corpus linguistic techniques for studying parliamentary discourse Extracting and interpreting frequency
Daily reflection

Day 5

Combining distant and close reading for studying parliamentary discourse
Concordance analysis
Daily reflection

Week 2 (15 hours + an open teaser session)