一句话：提出 ToxiShield，浏览器扩展实时识别、分类并去毒化开源代码评审中的毒性言论。
开源开发中的毒性互动会损害社区协作。论文提出 ToxiShield，一个实时浏览器扩展，用于识别并去毒化毒性代码评审。框架含毒性识别、带理由多类分类和去毒化三模块。在 38,761 条评审文本上，微调 BERT 二分类器达 97% F1；在 1,200 样本上，提示工程 Claude 3.5 Sonnet 多类分类达 39% MCC 与 42% F1；微调 Llama 3.2 去毒化模型获 95.27% 风格迁移准确率、97.03% 流畅度、67.07% 内容保留率和 84% J-score。10 名开发者验证表明其有助于营造更包容的开源环境。
Toxic interactions in open-source software development harm community collaboration. To combat this, we propose ToxiShield, a realtime browser extension that identifies and detoxifies toxic code reviews. The framework comprises three modules: toxicity identification, reasoned multiclass classification, and code review detoxification. Our fine-tuned BERT-based binary classifier achieved a 97% F1-score on 38,761 code review texts. For multiclass classification, Claude 3.5 Sonnet with prompt engineering achieved a 39% MCC and 42% F1 on 1,200 samples. Finally, our fine-tuned Llama 3.2 detoxification model reached 95.27% style transfer accuracy, 97.03% fluency, 67.07% content preservation, and an 84% J-score. Validation with 10 software developers suggests ToxiShield effectively fosters a more inclusive open-source environment.
DOI 原文 · arXiv · PDF（开放获取） · DBLP
@inproceedings{abs-2604-08886,
title = {Real-Time Toxicity Filtering for Open-Source Code Reviews},
author = {Md Awsaf Alam Anindya and Showvik Biswas and Anindya Iqbal and Jaydeb Sarker and Amiangshu Bosu},
booktitle = {FSE},
year = {2026},
doi = {10.1145/3803437.3807395}
}