Label-Efficient Self-Training for Attribute Extraction from Semi-Structured Web Documents
This addresses the costly and time-consuming labeling process for web document extraction, benefiting applications like knowledge base construction and personalized recommendation, with incremental improvements in efficiency and performance.
The paper tackles the problem of extracting structured information from HTML documents with limited human-labeled data, proposing LEAST, a label-efficient self-training method that uses a few human-labeled pages to pseudo-annotate unlabeled pages and trains a transferable model with re-weighting to mitigate noise. Experiments show it outperforms previous state-of-the-art by over 26 average F1 points on unseen websites, reducing the required human-labeled pages by more than 10x.
Extracting structured information from HTML documents is a long-studied problem with a broad range of applications, including knowledge base construction, faceted search, and personalized recommendation. Prior works rely on a few human-labeled web pages from each target website or thousands of human-labeled web pages from some seed websites to train a transferable extraction model that generalizes on unseen target websites. Noisy content, low site-level consistency, and lack of inter-annotator agreement make labeling web pages a time-consuming and expensive ordeal. We develop LEAST -- a Label-Efficient Self-Training method for Semi-Structured Web Documents to overcome these limitations. LEAST utilizes a few human-labeled pages to pseudo-annotate a large number of unlabeled web pages from the target vertical. It trains a transferable web-extraction model on both human-labeled and pseudo-labeled samples using self-training. To mitigate error propagation due to noisy training samples, LEAST re-weights each training sample based on its estimated label accuracy and incorporates it in training. To the best of our knowledge, this is the first work to propose end-to-end training for transferable web extraction models utilizing only a few human-labeled pages. Experiments on a large-scale public dataset show that using less than ten human-labeled pages from each seed website for training, a LEAST-trained model outperforms previous state-of-the-art by more than 26 average F1 points on unseen websites, reducing the number of human-labeled pages to achieve similar performance by more than 10x.