File size: 2,982 Bytes
9726ca4
 
25320a4
 
 
 
 
 
 
9726ca4
f06760c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
42
43
44
45
46
47
48
49
50
51
---
license: cc-by-nc-sa-4.0
language:
- de
tags:
- public participation
- text classification
- argument quality
- subjectivity
---

## Model Description

**bert-base-german-cased_cimt-argument-subjectivity-4** is a fine-tuned BERT model that is built to predict how subjective annotators perceive the concreteness of argumentative utterances in urban planning processes. It has been trained to recognize 4 classes: subjective (LABEL_3), rather subjective (LABEL_2), rather objective (LABEL_1), and objective (LABEL_0).

Specifically, this model is a *bert-base-german-cased* that was fine-tuned on the deviation of annotation decisions in https://github.com/juliaromberg/cimt-argument-concreteness-dataset.

## Background

This work is based on research in the project CIMT, which investigates the chances and challenges of involving citizens in political decisions in the context of sustainable mobility transitions. (for more information, visit https://www.cimt-hhu.de/en/)

## Details & Evaluation Results 

Can be found in the corresponding publication https://aclanthology.org/2022.argmining-1.11/.

## Usage
```python
from transformers import AutoModelForSequenceClassification, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("juliaromberg/bert-base-german-cased_cimt-argument-subjectivity-4")
model = AutoModelForSequenceClassification.from_pretrained("juliaromberg/bert-base-german-cased_cimt-argument-subjectivity-4")
```

### Citation
```
@inproceedings{romberg-2022-perspective,
    title = "Is Your Perspective Also My Perspective? Enriching Prediction with Subjectivity",
    author = "Romberg, Julia",
    editor = "Lapesa, Gabriella  and
      Schneider, Jodi  and
      Jo, Yohan  and
      Saha, Sougata",
    booktitle = "Proceedings of the 9th Workshop on Argument Mining",
    month = oct,
    year = "2022",
    address = "Online and in Gyeongju, Republic of Korea",
    publisher = "International Conference on Computational Linguistics",
    url = "https://aclanthology.org/2022.argmining-1.11",
    pages = "115--125",
    abstract = "Although argumentation can be highly subjective, the common practice with supervised machine learning is to construct and learn from an aggregated ground truth formed from individual judgments by majority voting, averaging, or adjudication. This approach leads to a neglect of individual, but potentially important perspectives and in many cases cannot do justice to the subjective character of the tasks. One solution to this shortcoming are multi-perspective approaches, which have received very little attention in the field of argument mining so far. In this work we present PerspectifyMe, a method to incorporate perspectivism by enriching a task with subjectivity information from the data annotation process. We exemplify our approach with the use case of classifying argument concreteness, and provide first promising results for the recently published CIMT PartEval Argument Concreteness Corpus.",
}
```