File size: 1,257 Bytes
937e7dc 016df69 937e7dc 016df69 937e7dc 016df69 937e7dc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 |
---
language: de
license: mit
datasets:
- Wahl-o-Mat
- Wahlprogramme
---
# GermanPolitical-Gelectra-base
Released May 2022, we fine-tuned the german electra model pre-trained by german-nlp-group (orginal model can be found at 'german-nlp-group/electra-base-german-uncased') to classify political sentiment of popular german parties.
Supported parties at the moment are "CDU/CSU", "SPD", "FDP", "Grüne", "Die Linke", "AfD".
## Overview
**Architecture:** Gelectra-base (uncased)
**Language:** German
## Datasets & Methodology
![Methodology for model]](https://www.wahl-o-bot.de/img/Method.f5dbe4fd.png)
For more information about data and methodology visit: https://www.wahl-o-bot.de/Methodik
## Production
The model can be used for free at https://www.wahl-o-bot.de
## Authors
Feel free to contact us about questions or feedback. We are always happy to help.
* Maximilian Witte: `maximilian.witte [at] uni-hamburg.de`
* Jasper Schwenzow: `jasper.schwenzow [at] uni-hamburg.de`
* Mark Heitmann: `mark.heitmann [at] uni-hamburg.de`
* Matthias Assenmacher: `[email protected] [at] muenchen.de`
The model is built in collaboration of the University of Hamburg and Ludwig Maximilians University Munich.
https://www.bwl.uni-hamburg.de/ci.html |