Detecting Contextomized Quotes in News Headlines by Contrastive Learning

Quotes are critical for establishing credibility in news articles. A direct quote enclosed in quotation marks has a strong visual appeal and is a sign of a reliable citation. Unfortunately, this journalistic practice is not strictly Figure 1: The central idea of QuoteCSE is based on followed, and a quote in the headline is often journalism principles, where quotes from news headlines "contextomized." Such a quote uses words out and body text should be matched. The proposed of context in a way that alters the speaker's contrastive learning framework maximizes the intention so that there is no semantically semantic similarity between the headline quote and the matching quote in the body text. We present matched quote in the body text while minimizing the QuoteCSE, a contrastive learning framework similarity for other unmatched quotes in the same or that represents the embedding of news quotes other articles.