This paper presents a novel approach for validating an established rule-based classification system of historical newspaper articles related to the “Great Moon Hoax of 1835” using event extraction techniques. The rule-based system classifies articles into five distinct categories: “Spreading,” “Debunking,” “Discussing,” “Recalling,” and “Mentioning,” based on a multidimensional framework incorporating lexical features, positional weighting, sentiment analysis, and temporal markers. The computed classification demonstrates around 95% match with the manual classification. To verify these classifications, I implement a complementary event extraction methodology that identifies event structures from historical texts. The verification framework applies hierarchical classification rules similar to the original system but derives its features from extracted events and their argumentative structures rather than lexical patterns alone. Applied to the corpus of 151 newspaper articles from the Library of Congress and British Newspaper Archive (1835–1900), this approach achieves around 86% agreement rate with the rule-based model, validating the robustness of the original classification framework. Analysis of disagreement patterns identifies specific challenges in distinguishing between “Discussing” and “Recalling” categories, where event extraction provides valuable contextual information not captured by lexical features alone. This dual-methodology approach demonstrates how event extraction can be effectively applied to historical texts to verify classification results.