Generative Models are Unsupervised Predictors of Page Quality: A Colossal-Scale Study
This addresses the challenge of identifying low-quality content at scale for web users and platforms, though it is incremental in applying existing models to a new task.
The study tackled the problem of detecting low-quality web content by showing that classifiers trained to discriminate human vs. machine-generated text can serve as unsupervised predictors of page quality, enabling fast bootstrapping in low-resource settings, and conducted analysis on 500 million web articles.
Large generative language models such as GPT-2 are well-known for their ability to generate text as well as their utility in supervised downstream tasks via fine-tuning. Our work is twofold: firstly we demonstrate via human evaluation that classifiers trained to discriminate between human and machine-generated text emerge as unsupervised predictors of "page quality", able to detect low quality content without any training. This enables fast bootstrapping of quality indicators in a low-resource setting. Secondly, curious to understand the prevalence and nature of low quality pages in the wild, we conduct extensive qualitative and quantitative analysis over 500 million web articles, making this the largest-scale study ever conducted on the topic.