Summary of Njust-kmg at Trac-2024 Tasks 1 and 2: Offline Harm Potential Identification, by Jingyuan Wang et al.
NJUST-KMG at TRAC-2024 Tasks 1 and 2: Offline Harm Potential Identification
by Jingyuan Wang, Shengdong Xu, Yang Yang
First submitted to arxiv on: 26 Mar 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG)
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 The proposed method, designed for TRAC-2024 Offline Harm Potential identification, consists of two sub-tasks: assessing harm likelihood and identifying potential targets. A rich dataset of social media comments in Indian languages, annotated by expert judges, was used to train algorithms capable of accurately predicting offline harm scenarios. The approach utilizes pre-trained models fine-tuned with contrastive learning techniques and an ensemble method for the test set. Notably, the proposed method ranked second in two separate tracks, achieving F1 values of 0.73 and 0.96 respectively. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A team of researchers developed a new way to identify situations where harm might occur offline, using comments from social media platforms in Indian languages. They trained special algorithms to predict the likelihood of harm and figure out who might be affected. The approach worked well, with the algorithm ranking second in two different competitions. This method could help prevent harmful situations by predicting where they might happen. |
Keywords
* Artificial intelligence * Likelihood
