一句话:比较 Stack Overflow 人类答案与 ChatGPT 生成答案,发现语义相似但人类答案整体质量高约 10%。
Stack Overflow 因生成式 AI 冲击出现流量下降,并在 ChatGPT 发布 6 天后因其答案质量低而禁止使用。本文通过自动比较与人工研究,系统评估人类撰写答案与 ChatGPT 生成答案的质量。结果显示,两类答案在语义上相似,但人类答案在多个维度上持续优于 ChatGPT,总体得分高出约 10%。作者公开了数据、分析脚本与详细结果,为软件问答社区是否应接纳生成式 AI 提供实证依据。
Stack Overflow, the world's largest software Q&A (SQA) website, is facing a significant traffic drop due to the emergence of generative AI techniques. ChatGPT is banned by Stack Overflow after only 6 days from its release. The main reason provided by the official Stack Overflow is that the answers generated by ChatGPT are of low quality. To verify this, we conduct a comparative evaluation of human-written and ChatGPT-generated answers. Our methodology employs both automatic comparison and a manual study. Our results suggest that human-written and ChatGPT-generated answers are semantically similar, however, human-written answers outperform ChatGPT-generated ones consistently across multiple aspects, specifically by 10% on the overall score. We release the data, analysis scripts, and detailed results at https://github.com/maxxbw54/GAI4SQA.
DOI 原文 · arXiv · PDF(开放获取) · DBLP
@inproceedings{XuNLHLKGNWLL23,
title = {Are We Ready to Embrace Generative AI for Software Q&A?},
author = {Bowen Xu and Thanh-Dat Nguyen and Thanh Le-Cong and Thong Hoang and Jiakun Liu and Kisub Kim and Chen Gong and Changan Niu and Chenyu Wang and Bach Le and David Lo},
booktitle = {ASE},
year = {2023},
doi = {10.1109/ase56229.2023.00023}
}