AMI @ EVALITA2020: Automatic Misogyny Identification

Fersini, Elisabetta; Nozza, Debora; Rosso, Paolo

doi:10.4000/books.aaccademia.6764

Cited by 62 publications

(38 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Encouragingly, more and more non-English datasets are being created, particularly for shared tasks (e.g. Wiegand et al, 2018;Ptaszynski et al, 2019;Fersini et al, 2020;Zampieri et al, 2020;Mulki and Ghanem, 2021). However, very few datasets cover more than one language (Ousidhoum et al, 2019;, and to our knowledge no dataset covers as many languages as MHC.…”

Section: Related Workmentioning

confidence: 99%

HateCheck: Functional Tests for Hate Speech Detection Models

Röttger¹,

Vidgen²,

Nguyen³

et al. 2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

102

View full text Add to dashboard Cite

Detecting online hate is a difficult task that even state-of-the-art models struggle with. Typically, hate speech detection models are evaluated by measuring their performance on held-out test data using metrics such as accuracy and F1 score. However, this approach makes it difficult to identify specific model weak points. It also risks overestimating generalisable model performance due to increasingly well-evidenced systematic gaps and biases in hate speech datasets. To enable more targeted diagnostic insights, we introduce HATECHECK, a suite of functional tests for hate speech detection models. We specify 29 model functionalities motivated by a review of previous research and a series of interviews with civil society stakeholders. We craft test cases for each functionality and validate their quality through a structured annotation process. To illustrate HATECHECK's utility, we test near-state-of-the-art transformer models as well as two popular commercial models, revealing critical model weaknesses.

show abstract

Section: Related Workmentioning

confidence: 99%

HateCheck: Functional Tests for Hate Speech Detection Models

Röttger¹,

Vidgen²,

Nguyen³

et al. 2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

102

View full text Add to dashboard Cite

show abstract

“…The related synthetic test set (Nozza et al, 2019) was created via several manually defined templates and synonyms for "woman" as identity terms. MISOGYNY (ITA) (Fersini et al, 2020) is the benchmark corpus for misogyny detection in Italian. The synthetic test set has been generated similarly to the English one.…”

Section: Experimental Settingsmentioning

confidence: 99%

Entropy-based Attention Regularization Frees Unintended Bias Mitigation from Lists

Attanasio¹,

Nozza²,

Hovy³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Warning: This paper contains examples of language that some people may find offensive.Natural Language Processing (NLP) models risk overfitting to specific terms in the training data, thereby reducing their performance, fairness, and generalizability. E.g., neural hate speech detection models are strongly influenced by identity terms like gay, or women, resulting in false positives, severe unintended bias, and lower performance. Most mitigation techniques use lists of identity terms or samples from the target domain during training. However, this approach requires a-priori knowledge and introduces further bias if important terms are neglected. Instead, we propose a knowledge-free Entropy-based Attention Regularization (EAR) to discourage overfitting to training-specific terms. An additional objective function penalizes tokens with low self-attention entropy. We fine-tune BERT via EAR: the resulting model matches or exceeds state-of-the-art performance for hate speech classification and bias metrics on three benchmark corpora in English and Italian. EAR also reveals overfitting terms, i.e., terms most likely to induce bias, to help identify their effect on the model, task, and predictions.

show abstract

“…In the last years, many shared tasks have been organized with the aim of detecting hate speech on social media comments (Kumar et al, 2018;Basile et al, 2019;Zampieri et al, 2020, inter alia). While the majority of them focus on English, some efforts have been made to include other languages (e.g., Italian, Arabic) (Bosco et al, 2018;Fersini et al, 2018;Wiegand et al, 2018;Fersini et al, 2020b;Mulki and Ghanem, 2021, inter alia). Chaudhary et al (2021) proposed a one-of-a-kind shared task for Homophobia and Transphobia detection on social comments for three languages (English, Tamil, and code-mixed Tamil-English).…”

Section: Related Workmentioning

confidence: 99%

Nozza@LT-EDI-ACL2022: Ensemble Modeling for Homophobia and Transphobia Detection

Nozza¹

2022

Proceedings of the Second Workshop on Language Technology for Equality, Diversity and Inclusion

Self Cite

View full text Add to dashboard Cite

In this paper, we describe our approach for the task of homophobia and transphobia detection in English social media comments. The dataset consists of YouTube comments, and it has been released for the shared task on Homophobia/Transphobia Detection in social media comments. Given the high class imbalance, we propose a solution based on data augmentation and ensemble modeling. We fine-tuned different large language models (BERT, RoBERTa, and HateBERT) and used the weighted majority vote on their predictions. Our proposed model obtained 0.48 and 0.94 for macro and weighted F1-score, respectively, ranking at the third position.

show abstract

AMI @ EVALITA2020: Automatic Misogyny Identification

Cited by 62 publications

References 39 publications

HateCheck: Functional Tests for Hate Speech Detection Models

HateCheck: Functional Tests for Hate Speech Detection Models

Entropy-based Attention Regularization Frees Unintended Bias Mitigation from Lists

Nozza@LT-EDI-ACL2022: Ensemble Modeling for Homophobia and Transphobia Detection

Contact Info

Product

Resources

About