Commit
•
9dfb86d
1
Parent(s):
75a7303
Added general information
Browse files
README.md
CHANGED
|
@@ -1,3 +1,25 @@
|
|
| 1 |
---
|
| 2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- de
|
| 5 |
+
tags:
|
| 6 |
+
- text-classication
|
| 7 |
+
metrics:
|
| 8 |
+
- precision
|
| 9 |
+
widget:
|
| 10 |
+
- text: "Ein Blick auf die Karte der Ertragslagen von Windkraftanlagen im Bayerischen Windatlas zeigt klar, dass gerade das südliche Allgäu Windkraftanlagen wenig Sinn machen. Leider fehlt im Artikel ain Verweis auf deises leicht nachprüfbares Faktum."
|
| 11 |
+
example_classification: "Enthält Ansprache an Redaktion"
|
| 12 |
---
|
| 13 |
+
|
| 14 |
+
# WTWM gpt2 based mentions detector
|
| 15 |
+
|
| 16 |
+
This is a model for the task of classifying whether or not a articles comment addresses the moderation team/authors of the media house that published the article. In this prototype stage the media houses are Bayerischer Rundfunk and Mitteldeutscher Rundfunk.
|
| 17 |
+
|
| 18 |
+
This classification task is implemented as a binary classification into:
|
| 19 |
+
|
| 20 |
+
* label 0: the comment holds no mention
|
| 21 |
+
* label 1: the comment addresses the moderation team/authors of the media house
|
| 22 |
+
|
| 23 |
+
For this task, we decided to use [german-gpt2](https://huggingface.co/dbmdz/german-gpt2) by MDZ of Bayerische Staatsbibliothek as a foundation model.
|
| 24 |
+
|
| 25 |
+
**Our finetuned model is still work in progress. More information to come.**
|