File size: 993 Bytes
d8c2115
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
---
tags:
- subjectivity
- newspapers
- CLEF2023
---

Fine-tuned [mDeBERTa V3](https://huggingface.co/microsoft/mdeberta-v3-base) model for subjectivity detection in newspaper sentences.
This model was developed as part of the CLEF 2023 CheckThat! Lab [Task 2: Subjectivity in News Articles](https://checkthat.gitlab.io/clef2023/task2/).

The goal in this task is to detect whether a sentence is objective (OBJ) or subjective (SUBJ). A sentence is subjective if its content is based on or influenced by personal feelings, tastes, or
opinions. Otherwise, the sentence is objective. [(Antici et al., 2023)](https://ceur-ws.org/Vol-3370/paper10.pdf).

The model was fine-tuned using a multilingual training and development dataset, for which the following (hyper)parameters were utilized:
```
Batch Size    = 64
Max Epochs    = 8
Learning Rate = 3e-5
Warmup Steps  = 500
Weight Decay  = 0.3
```

The model ranked second in the CheckThat! Lab and obtained a macro F1 of 0.81 and a SUBJ F1 of 0.81.