OLEPA is a semisupervised information-extraction system that produces extraction rules from semistructured Web documents without requiring detailed annotation of the training documents. It performs well for program-generated Web pages with few training pa