Summary of Pejorativity: Disambiguating Pejorative Epithets to Improve Misogyny Detection in Italian Tweets, by Arianna Muti et al.
PejorativITy: Disambiguating Pejorative Epithets to Improve Misogyny Detection in Italian Tweets
by Arianna Muti, Federico Ruggeri, Cagri Toraman, Lorenzo Musetti, Samuel Algherini, Silvia Ronchi, Gianmarco Saretto, Caterina Zapparoli, Alberto Barrón-Cedeño
First submitted to arxiv on: 3 Apr 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents PejorativITy, a novel corpus of 1,200 manually annotated Italian tweets that can help detect misogynistic language. The corpus is designed to identify both pejorative words and sentences with negative connotations. To improve the accuracy of misogyny detection models, the authors propose injecting information about disambiguated words into these models. They explore two approaches for doing so: concatenating pejorative information or substituting ambiguous words with unambiguous terms. Experimental results show that both methods lead to significant improvements in classification accuracy, suggesting that word sense disambiguation is a valuable preliminary step in detecting misogyny. The authors also analyze the understanding of pejorative epithets by large language models (LLMs) using contextual word embeddings analysis and prompting. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine if you could identify when someone is saying something mean or unfair just because they’re talking about women. This paper helps create a system that can do exactly that. They made a special collection of tweets where words are labeled as either nice or mean, so computers can learn to recognize misogynistic language. The authors tested their idea by adding information about these “mean” words into a computer program designed to detect misogyny. Surprisingly, this worked really well! They also looked at how large language models understand the meaning of these words. This research is important because it could help us create more accurate systems for detecting and preventing online harassment. |
Keywords
» Artificial intelligence » Classification » Prompting
