UniBO @ AMI: A Multi-Class Approach to Misogyny and Aggressiveness Identification on Twitter Posts Using AlBERTo

Muti, Arianna; Barrón-Cedeño, Alberto

doi:10.4000/books.aaccademia.6769

Cited by 7 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Abusive Language. Typically, abusive language refers to a wide range of concepts (Balayn et al, 2021;Poletto et al, 2021), including hate speech (Yin and Zubiaga, 2021;Alkomah and Ma, 2022;Jain and Sharma, 2022), profanity (Soykan et al, 2022), aggressive language (Muti et al, 2022;Kanclerz et al, 2021), offensive language (Pradhan et al, 2020;Kogilavani et al, 2021), cyberbullying (Rosa et al, 2019) and misogyny (Shushkevich and Cardiff, 2019). Pamungkas et al (2023) overview recent research across domains and languages.…”

Section: Background and Related Workmentioning

confidence: 99%

Vicinal Risk Minimization for Few-Shot Cross-lingual Transfer in Abusive Language Detection

De la Peña Sarracén,

Rosso,

Litschko

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

Cross-lingual transfer learning from highresource to medium and low-resource languages has shown encouraging results. However, the scarcity of resources in target languages remains a challenge. In this work, we resort to data augmentation and continual pre-training for domain adaptation to improve cross-lingual abusive language detection. For data augmentation, we analyze two existing techniques based on vicinal risk minimization and propose MIXAG, a novel data augmentation method which interpolates pairs of instances based on the angle of their representations. Our experiments involve seven languages typologically distinct from English and three different domains. The results reveal that the data augmentation strategies can enhance fewshot cross-lingual abusive language detection. Specifically, we observe that consistently in all target languages, MIXAG improves significantly in multidomain and multilingual environments. Finally, we show through an error analysis how the domain adaptation can favour the class of abusive texts (reducing false negatives), but at the same time, declines the precision of the abusive language detection model.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

Vicinal Risk Minimization for Few-Shot Cross-lingual Transfer in Abusive Language Detection

De la Peña Sarracén,

Rosso,

Litschko

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

show abstract

“…For Task antiLGBT , we adopt a multi-class approach with mutually exclusive categories with three output units. This approach is based on the top-performing model (Muti and Barrón-Cedeño, 2020) at the AMI shared task on the identification of misogynous and aggressive tweets (Elisabetta Fersini, 2020). No external data is considered in this model.…”

Section: Task Antilgbtmentioning

confidence: 99%

LeaningTower@LT-EDI-ACL2022: When Hope and Hate Collide

Muti¹,

Manerba²,

Korre³

et al. 2022

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

Self Cite

View full text Add to dashboard Cite

The 2022 edition of LT-EDI proposed two tasks in various languages. Task hope required models for the automatic identification of hopeful comments for equality, diversity, and inclusion. Task antiLGBT focused on the identification of homophobic and transphobic comments. We targeted both tasks in English by using reinforced BERT-based approaches. Our core strategy aimed at exploiting the data available for each given task to augment the amount of supervised instances in the other. On the basis of an active learning process, we trained a model on the dataset for Task i and applied it to the dataset for Task j to iteratively integrate new silver data for Task i. Our official submissions to the shared task obtained a macro-averaged F 1 score of 0.53 for Task hope and 0.46 for Task antiLGBT , placing our team in the third and fourth positions out of 11 and 12 participating teams respectively.

show abstract

“…The output for each classifier is a sigmoid function too. We opt for this approach after observing that treating the classes separately increased the performance in a multi-class model predicting misogynous, misogynous-aggressive or none (Muti and Barrón-Cedeño, 2020). This approach allows us to predict multiple mutually non-exclusive classes.…”

Section: System Overviewmentioning

confidence: 99%

“…The task is more challenging than when dealing with text alone because, in general, both the textual and the visual channels play an indivisible role in conveying the desired message. 1 We build upon our previous experience in identifying misogyny and aggressiveness in text (Muti and Barrón-Cedeño, 2020) and approach both multimodal tasks with a supervised multi-modal bitransformer model (MMBT) (Kiela et al, 2020a). We use bert-base-uncased-hatexplain (Mathew et al, 2020) and bert-base-uncased (Devlin et al, 2019) for the textual embeddings, and CLIP (Radford et al, 2021) for the visual ones.…”

Section: Introductionmentioning

confidence: 99%

UniBO at SemEval-2022 Task 5: A Multimodal bi-Transformer Approach to the Binary and Fine-grained Identification of Misogyny in Memes

Muti¹,

Korre²,

Barrón-Cedeño³

2022

Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)

Self Cite

View full text Add to dashboard Cite

We present our submission to SemEval 2022 Task 5 on Multimedia Automatic Misogyny Identification. We address the two tasks: Task A consists of identifying whether a meme is misogynous. If so, Task B attempts to identify its kind among shaming, stereotyping, objectification, and violence. Our approach combines a BERT Transformer with CLIP for the textual and visual representations. Both textual and visual encoders are fused in an early-fusion fashion through a Multimodal Bidirectional Transformer with unimodally pretrained components. Our official submissions obtain macroaveraged F 1 =0.727 in Task A (4th position out of 69 participants) and weighted F 1 =0.710 in Task B (4th position out of 42 participants).

show abstract

UniBO @ AMI: A Multi-Class Approach to Misogyny and Aggressiveness Identification on Twitter Posts Using AlBERTo

Cited by 7 publications

References 12 publications

Vicinal Risk Minimization for Few-Shot Cross-lingual Transfer in Abusive Language Detection

Vicinal Risk Minimization for Few-Shot Cross-lingual Transfer in Abusive Language Detection

LeaningTower@LT-EDI-ACL2022: When Hope and Hate Collide

UniBO at SemEval-2022 Task 5: A Multimodal bi-Transformer Approach to the Binary and Fine-grained Identification of Misogyny in Memes

Contact Info

Product

Resources

About