ParlaCAP – Comparing agenda settings across parliaments via the ParlaMint dataset

Project factsheet

Project summary

The ParlaCAP project leverages advanced natural language processing to analyse political agendas and sentiments in debates from European national and regional parliaments. The automatic coding of agendas throughout a wide dataset of more than 8 million speeches, given in more than 20 languages, has become possible recently with significant developments in natural language processing and artificial intelligence, allowing for multilingual transformer models to provide both highly consistent and accurate codings. By integrating the ParlaMint dataset and the Comparative Agendas Project’s coding scheme, the project creates a comprehensive, FAIR dataset for comparative political research, enhancing transparency and accountability in legislative discourse across Europe.

Challenge

Parliaments are the cornerstone of democracy in Europe, ensuring the political representation of citizens. Despite their empirical relevance, parliamentary studies have often limited their scope to a single parliamentary body or a small group of parliaments analysed in comparative perspective.

The main challenge of the ParlaCAP project is to bridge the gap between existing parliamentary research data and how to utilise these data in political science research by integrating two key international and cross-disciplinary initiatives: the CLARIN ERIC ParlaMint project, which provides texts of parliamentary debates from European national parliaments, and the Comparative Agendas Project (CAP), which offers a coding schema of 21 topics for tracking political agendas in parliamentary proceedings.

Solution

The project employs the Comparative Agendas Project’s text-as-data methodology to analyse parliamentary debates, by automatically coding the agenda of each speech and transforming the ParlaMint corpora into a structured and tabular dataset, available for complete download through CESSDA ERIC.

Each speech is further coded with the sentiment expressed, and the data are cross-referenced with the PartyFacts metadatabase on political party metadata and the V-DEM surveys on the state of democracies. With this enriched and fully FAIR dataset, suitable for quantitative research, it is possible to acquire a comprehensive understanding of how political attention is distributed across policy areas by analysing topic and sentiment coding over an unprecedented number of parliaments. The dataset is available through research infrastructures such as CESSDA, CLARIN and DARIAH, along with a graphical user interface and an API for broader accessibility.

Results

• ParlaCAP dataset v1.0 – 8 million speeches from 28 European national and regional parliaments, each coded with the sentiment expressed (ParlaSent coding from negative, over neutral, to positive) and the topic discussed (CAP coding with 21 topics), with rich metadata on the speakers, parties and democracies. The dataset extends ParlaMint 5.0 and simplifies the data to a tabular form. Repository
• ParlaCAP web interface – lets users explore and download the dataset without writing any code: speeches can be queried by parliament, speaker, party, topic, sentiment, date range and more, and accessed programmatically through an API. parlacap.ipipan.waw.pl
• ParlaCAP topic classifier – a multilingual model assigning CAP topic categories to parliamentary speeches, built by fine-tuning XLM-R-Parla on GPT-4o-annotated debates from multiple European parliaments. Model on Hugging Face · Guide for automatic CAP annotation of ParlaMint data
• ParlaSent model – a multilingual transformer model for sentiment analysis in parliamentary speeches. Model on Hugging Face
• ParlaMint 5.0 – comparable corpora of parliamentary debates from 29 European countries and regions, annotated with CAP topics and sentiment. Repository
• Analysis tutorials – step-by-step Jupyter notebooks in Python and R covering sentiment analysis, party comparisons and cross-country analyses. Tutorials

More information about the project results is available on the OSCARS project page.