From Text to Transparency: Mapping Political Advocacy and Stances
| Project Title: From Text to Transparency: Mapping Political Advocacy and Stances | |
| Funded by: ETHZ, Swiss Data Science Center (SDSC) | |
| Duration: 2025 - 2027 | |
| Principal Investigators: Dr. Laurence Brandenberger (IPZ, UZH), Dr. Sophia Schlosser (IPZ, UZH) | |
| Amount: 242100 CHF | |
| Focus: Mapping political advocacy, detecting stances, and improving political transparency | |
| Data: DemocraSci Knowledge Graph | |
| Methods: Argument Mining, Stance Detection, Embedding-Based Topic Modeling, Multilingual NLP | |
| Open Access: Advocacy maps, stance models, and enhanced embed2discover tool will be made publicly available | |
| Tagline: "Showing What Politicians Stand For: Transparency through Data" | |
| Acronym: MappingDemocraSci |
About the Project
In today’s democracies, the lack of accessible, neutral information about elected representatives‘ true positions weakens the bond between voters and politicians.
MappingDemocraSci tackles this issue by analyzing Swiss parliamentary bills and speeches to map the advocacy efforts and political stances of Members of Parliament (MPs).
Leveraging cutting-edge natural language processing (NLP) methods, we identify what politicians advocate for, how they take stances, and how consistently they align with their declared interests.
Learn More
Key Objectives:
- Advance Argument Detection: Developing tools to automatically detect advocacy arguments in political texts.
- Stance Detection: Building multilingual models to detect whether MPs favor or oppose specific issues.
- Advocacy Mapping: Constructing detailed maps of the issues and interests MPs support — across hundreds of policy topics.
- Stance Congruence Analysis: Studying how consistent MPs‘ stances are and how social networks shape decision-making.
Impact:
- For Researchers: New semi-automated methods to discover political advocacy patterns.
- For Citizens: Clear indices showing what their elected representatives stand for.
- For Democracy: A major step toward transparency, accountability, and informed voting.
Integration:
The project’s outputs will be incorporated into an upgraded version of the embed2discover web tool — and will be made available to the public via the DemocraSci website and the other political platforms.