Skip to content

README.md

Latest commit

 

History

History
275 lines (209 loc) · 11.4 KB

README.md

File metadata and controls

275 lines (209 loc) · 11.4 KB

SuperMAMI Logo

SuperMAMI

Synthesizing Up-to-date Publications & bEnchmarks for Researchers
in Multimedia Automatic Misogyny Identification

A central hub for our research, papers, datasets, and resources on misogyny detection in online multimedia content.

📑 Table of Contents

💡 About

SUPER MAMI is a dissemination hub for our work on detecting and understanding misogyny in multimedia content.
We focus on:

  • 🕵️‍♀️ Detection of misogynistic language and content
  • 📚 Understanding patterns across platforms, languages, and modalities
  • 🤝 Providing datasets, benchmarks, and papers for the research community

Our goal is to advance knowledge, support reproducibility, and foster collaboration in the study of online misogyny.

📚 Publications

This collection showcases our ongoing research into the pervasive issues of sexism and misogyny, particularly within digital and social contexts. Our work spans various facets, including:

🎯 Detection and Classification: Developing and refining computational models to identify and categorize both unimodal and multimodal misogynistic content.

🎯 Bias Estimation: Measuring and quantifying biases in language models and (unimodal and multimodal) datasets related to gender, sexism, and misogyny.

🎯 Bias Mitigation: Designing approaches to reduce bias and improve fairness in automated systems for detecting misogynistic content.

Papers

📚 "Benchmark dataset of memes with text transcriptions for automatic detection of multimodal misogynistic content" 
    ```
      @article{gasparini2022benchmark,
    title={Benchmark dataset of memes with text transcriptions for automatic detection of multi-modal misogynistic content},
    author={Gasparini, Francesca and Rizzi, Giulia and Saibene, Aurora and Fersini, Elisabetta},
    journal={Data in brief},
    volume={44},
    pages={108526},
    year={2022},
    publisher={Elsevier}
    }
    ```
📚 "Misogynous meme recognition: A preliminary study" 
    ```
      @inproceedings{fersini2021misogynous,
    title={Misogynous meme recognition: A preliminary study},
    author={Fersini, Elisabetta and Rizzi, Giulia and Saibene, Aurora and Gasparini, Francesca},
    booktitle={International conference of the Italian association for artificial intelligence},
    pages={279--293},
    year={2021},
    organization={Springer}
    }
    ```
📚 "SemEval-2022 Task 5: Multimedia automatic misogyny identification" 
    ```
      @inproceedings{fersini2022semeval,
    title={SemEval-2022 Task 5: Multimedia automatic misogyny identification},
    author={Fersini, Elisabetta and Gasparini, Francesca and Rizzi, Giulia and Saibene, Aurora and Chulvi, Berta and Rosso, Paolo and Lees, Alyssa and Sorensen, Jeffrey},
    booktitle={Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022">},
    pages={533--549},
    year={2022}
    }
    ```
📚 "Bias mitigation in misogynous meme recognition: A preliminary study" 
    ```
        @inproceedings{balducci2023bias,
    title={Bias Mitigation in Misogynous Meme Recognition: A Preliminary Study},
    author={Balducci, Gianmaria and Rizzi, Giulia and Fersini, Elisabetta},
    booktitle={Proceedings of the 9th Italian Conference on Computational Linguistics (CLiC-it 2023">},
    pages={63--69},
    year={2023}
    }
    ```
📚 "Recognizing misogynous memes: Biased models and tricky archetypes" GitHub 
    ```
        @article{rizzi2023recognizing,
    title={Recognizing misogynous memes: Biased models and tricky archetypes},
    author={Rizzi, Giulia and Gasparini, Francesca and Saibene, Aurora and Rosso, Paolo and Fersini, Elisabetta},
    journal={Information Processing \& Management},
    volume={60},
    number={5},
    pages={103474},
    year={2023},
    publisher={Elsevier}
    }
    ```
📚 "Multimodal Hate Speech Detection in Memes from Mexico using BLIP" 
    ```
       @article{maqbool2024multimodal,
   title={Multimodal Hate Speech Detection in Memes from Mexico using BLIP},
   author={Maqbool, Fariha and Fersini, Elisabetta},
   year={2024}
   }
    ```
📚 "A contrastive learning based approach to detect sexism in memes" 
    ```
       @article{maqbool2024contrastive,
    title={A contrastive learning based approach to detect sexism in memes},
    author={Maqbool, Fariha and Fersini, Elisabetta},
    journal={Working Notes of CLEF},
    year={2024}
    }
    ```
📚 "From Explanation to Detection: Multimodal Insights into Disagreement in Misogynous Memes" 
    ```
       @inproceedings{rizzi2024explanation,
    title={From Explanation to Detection: Multimodal Insights into Disagreement in Misogynous Memes},
    author={Rizzi, Giulia and Rosso, Paolo and Fersini, Elisabetta},
    booktitle={Proceedings of the 10th Italian Conference on Computational Linguistics (CLiC-it 2024">}},
    pages={821--828},
    year={2024}
    }
    ```
📚 "Misogynous Memes Recognition: Training vs Inference Bias Mitigation Strategies" 
    ```
       @article{balducci2025misogynous,
     title={Misogynous Memes Recognition: Training vs Inference Bias Mitigation Strategies},
     author={Balducci, Gianmaria and Rizzi, Giulia and Fersini, Elisabetta},
     journal={IJCoL. Italian Journal of Computational Linguistics},
     volume={11},
     number={11-1},
     year={2025},
     publisher={Accademia University Press}
     }
    ```

Abstracts

📚 "Misogynous Memes Recognition: Training vs Inference Bias Mitigation Strategies" 
    ```
        @article{rizzibeyond,
    title={Beyond Misogyny Detection: Investigating Bias and Embracing Perspectivism},
    author={Rizzi, Giulia and Fersini, Elisabetta},
    journal={Book of Abstracts _ Data Science & Social Research (DSSR 2025)},
    pages={13}
    }
    ```

Shared Tasks Participation

Our Participation in EXIST Shared Tasks :

📚 "AI-UPV at EXIST 2023--Sexism Characterization Using Large Language Models Under The Learning with Disagreements Regime" GitHub 
    ```
       @inproceedings{de2023ai,
    title={AI-UPV at EXIST 2023--Sexism Characterization Using Large Language Models Under The Learning with Disagreements Regime},
    author={de Paula, A and Rizzi, G and Fersini, E and Spina, D and others},
    booktitle={CEUR WORKSHOP PROCEEDINGS},
    volume={3497},
    pages={985--999},
    year={2023},
    organization={CEUR-WS}
    }
    ```
📚 "PINK at EXIST2024: a cross-lingual and multi-modal transformer approach for sexism detection in memes" GitHub 
    ```
      @@article{rizzi2024pink,
    title={PINK at EXIST2024: a cross-lingual and multi-modal transformer approach for sexism detection in memes},
    author={Rizzi, Giulia and Gimeno-G{\'o}mez, David and Fersini, Elisabetta and Mart{\'\i}nez-Hinarejos, Carlos-D},
    journal={Working Notes of CLEF},
    year={2024}
    }
    ```

📊 Datasets

Our datasets may be distributed upon request and for academic purposes only. To request the datasets, please fill out the respective forms:

Name Language Type of Data Number of Data Task Annotators Perspectivist Evaluation Additional Info
Benchmark 🇬🇧 MEME 800 A: Misogyny detection and B: Misogyny type 👥 crowd and 🎓 Domain Experts ❌ ❌
MAMI 🇬🇧 MEME 10k + 1k A: Misogyny detection and B: Misogyny type 👥 crowd ❌ ❌
MAMITA 🇮🇹 MEME 1880 A: Misogyny detection and B: Misogyny type 👥 crowd and 🎓 Domain Experts ✅ Demographic and socio-cultural background

👩‍🔬 Team

Created with ❤️ by the MAMI Research Team

⭐ Cite & Support

If you use our datasets or papers, please cite our work.
Don’t forget to star this repository to support our research!

GitHub Stars