COLD: A Benchmark for Chinese Offensive Language Detection
This work addresses the problem of offensive language detection for Chinese social media and AI models, but it is incremental as it builds on existing detection tasks by focusing on a specific language.
The authors tackled the lack of reliable datasets for Chinese offensive language detection by proposing the COLD benchmark, which includes a dataset and baseline detector, and they found that existing generative models expose offensive issues, with anti-bias content and specific keywords triggering such outputs more easily.
Offensive language detection is increasingly crucial for maintaining a civilized social media platform and deploying pre-trained language models. However, this task in Chinese is still under exploration due to the scarcity of reliable datasets. To this end, we propose a benchmark --COLD for Chinese offensive language analysis, including a Chinese Offensive Language Dataset --COLDATASET and a baseline detector --COLDETECTOR which is trained on the dataset. We show that the COLD benchmark contributes to Chinese offensive language detection which is challenging for existing resources. We then deploy the COLDETECTOR and conduct detailed analyses on popular Chinese pre-trained language models. We first analyze the offensiveness of existing generative models and show that these models inevitably expose varying degrees of offensive issues. Furthermore, we investigate the factors that influence the offensive generations, and we find that anti-bias contents and keywords referring to certain groups or revealing negative attitudes trigger offensive outputs easier.