MuTox: Universal MUltilingual Audio-based TOXicity Dataset and Zero-shot Detector
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910503378354176 |
|---|---|
| author | Costa-jussà, Marta R. Meglioli, Mariano Coria Andrews, Pierre Dale, David Hansanti, Prangthip Kalbassi, Elahe Mourachko, Alex Ropers, Christophe Wood, Carleigh |
| author_facet | Costa-jussà, Marta R. Meglioli, Mariano Coria Andrews, Pierre Dale, David Hansanti, Prangthip Kalbassi, Elahe Mourachko, Alex Ropers, Christophe Wood, Carleigh |
| contents | Research in toxicity detection in natural language processing for the speech modality (audio-based) is quite limited, particularly for languages other than English. To address these limitations and lay the groundwork for truly multilingual audio-based toxicity detection, we introduce MuTox, the first highly multilingual audio-based dataset with toxicity labels. The dataset comprises 20,000 audio utterances for English and Spanish, and 4,000 for the other 19 languages. To demonstrate the quality of this dataset, we trained the MuTox audio-based toxicity classifier, which enables zero-shot toxicity detection across a wide range of languages. This classifier outperforms existing text-based trainable classifiers by more than 1% AUC, while expanding the language coverage more than tenfold. When compared to a wordlist-based classifier that covers a similar number of languages, MuTox improves precision and recall by approximately 2.5 times. This significant improvement underscores the potential of MuTox in advancing the field of audio-based toxicity detection. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_05060 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | MuTox: Universal MUltilingual Audio-based TOXicity Dataset and Zero-shot Detector Costa-jussà, Marta R. Meglioli, Mariano Coria Andrews, Pierre Dale, David Hansanti, Prangthip Kalbassi, Elahe Mourachko, Alex Ropers, Christophe Wood, Carleigh Sound Computation and Language Audio and Speech Processing I.2.7 Research in toxicity detection in natural language processing for the speech modality (audio-based) is quite limited, particularly for languages other than English. To address these limitations and lay the groundwork for truly multilingual audio-based toxicity detection, we introduce MuTox, the first highly multilingual audio-based dataset with toxicity labels. The dataset comprises 20,000 audio utterances for English and Spanish, and 4,000 for the other 19 languages. To demonstrate the quality of this dataset, we trained the MuTox audio-based toxicity classifier, which enables zero-shot toxicity detection across a wide range of languages. This classifier outperforms existing text-based trainable classifiers by more than 1% AUC, while expanding the language coverage more than tenfold. When compared to a wordlist-based classifier that covers a similar number of languages, MuTox improves precision and recall by approximately 2.5 times. This significant improvement underscores the potential of MuTox in advancing the field of audio-based toxicity detection. |
| title | MuTox: Universal MUltilingual Audio-based TOXicity Dataset and Zero-shot Detector |
| topic | Sound Computation and Language Audio and Speech Processing I.2.7 |
| url | https://arxiv.org/abs/2401.05060 |