UniBoe’s at SemEval-2023 Task 10: Model-Agnostic Strategies for the Improvement of Hate-Tuned and Generative Models in the Classification of Sexist Posts

Muti, Arianna; Fernicola, Francesco; Barrón-Cedeño, Alberto

doi:10.18653/v1/2023.semeval-1.158

Cited by 2 publications

(2 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generating data: With the problem of data scarcity, especially in multilingual settings, some research studies are directed into providing more efficient solutions for data augmentation, by leveraging generated samples in order to gradually train their detection models and enhance the performance of their classification capabilities. Several approaches have been already released on English samples, that may be used to work on generating multilingual data, using different methods like adversarial auto-regressive models ( Ocampo, Cabrio & Villata, 2023 ), generative GPT3 PLM-based models ( Hartvigsen et al, 2022 ), or generative GPT-Neo based model ( Muti, Fernicola & Barrón-Cedeño, 2023 ).…”

Section: Challenges and Limitationsmentioning

confidence: 99%

A survey on multi-lingual offensive language detection

Mnassri,

Farahbakhsh,

Chalehchaleh

et al. 2024

PeerJ Computer Science

View full text Add to dashboard Cite

The prevalence of offensive content on online communication and social media platforms is growing more and more common, which makes its detection difficult, especially in multilingual settings. The term “Offensive Language” encompasses a wide range of expressions, including various forms of hate speech and aggressive content. Therefore, exploring multilingual offensive content, that goes beyond a single language, focus and represents more linguistic diversities and cultural factors. By exploring multilingual offensive content, we can broaden our understanding and effectively combat the widespread global impact of offensive language. This survey examines the existing state of multilingual offensive language detection, including a comprehensive analysis on previous multilingual approaches, and existing datasets, as well as provides resources in the field. We also explore the related community challenges on this task, which include technical, cultural, and linguistic ones, as well as their limitations. Furthermore, in this survey we propose several potential future directions toward more efficient solutions for multilingual offensive language detection, enabling safer digital communication environment worldwide.

show abstract

Section: Challenges and Limitationsmentioning

confidence: 99%

A survey on multi-lingual offensive language detection

Mnassri,

Farahbakhsh,

Chalehchaleh

et al. 2024

PeerJ Computer Science

View full text Add to dashboard Cite

show abstract

“…Curation of datasets [5,16,35] can also aid in sexism detection. With advancements made in Deep Learning (DL), especially after the introduction of transformer architecture [34], models like BERT [9] or RoBERTa [19] have become de-facto models that have been applied to detect sexism from text data [13,23,29]. Even though the aforementioned publications use the whole dataset to train and evaluate their models, some researchers [2-4, 12, 22] suggest that some data instances are more useful for driving the learning process and impacting the final model performance than others.…”

Section: Introductionmentioning

confidence: 99%

Sexism Detection on a Data Diet

Bandyopadhyay,

Assenmacher,

Alonso-Moral

et al. 2024

Companion Proceedings of the 16th ACM Web Science Conference

View full text Add to dashboard Cite

There is an increase in the proliferation of online hate commensurate with the rise in the usage of social media. In response, there is also a significant advancement in the creation of automated tools aimed at identifying harmful text content using approaches grounded in Natural Language Processing and Deep Learning. Although it is known that training Deep Learning models require a substantial amount of annotated data, recent line of work suggests that models trained on specific subsets of the data still retain performance comparable to the model that was trained on the full dataset. In this work, we show how we can leverage influence scores to estimate the importance of a data point while training a model and designing a pruning strategy applied to the case of sexism detection. We evaluate the model performance trained on data pruned with different pruning strategies on three out-of-domain datasets and find, that in accordance with other work a large fraction of instances can be removed without significant performance drop. However, we also discover that the strategies for pruning data, previously successful in Natural Language Inference tasks, do not readily apply to the detection of harmful content and instead amplify the already prevalent class imbalance even more, leading in the worst-case to a complete absence of the hateful class.Warning: This paper contains instances of hateful and sexist language to serve as examples.

show abstract

UniBoe’s at SemEval-2023 Task 10: Model-Agnostic Strategies for the Improvement of Hate-Tuned and Generative Models in the Classification of Sexist Posts

Cited by 2 publications

References 25 publications

A survey on multi-lingual offensive language detection

A survey on multi-lingual offensive language detection

Sexism Detection on a Data Diet

Contact Info

Product

Resources

About