Humor in Collective Discourse: Unsupervised Funniness Detection in the New Yorker Cartoon Caption Contest
This work addresses humor detection for improving conversational agents and text/multimodal systems, but it is incremental as it builds on existing sentiment and centrality methods.
The study tackled the problem of automatically detecting the funniest caption in the New Yorker Cartoon Caption Contest by comparing a dozen methods, finding that negative sentiment, human-centeredness, and lexical centrality most strongly predict humor, with positive sentiment also contributing.
The New Yorker publishes a weekly captionless cartoon. More than 5,000 readers submit captions for it. The editors select three of them and ask the readers to pick the funniest one. We describe an experiment that compares a dozen automatic methods for selecting the funniest caption. We show that negative sentiment, human-centeredness, and lexical centrality most strongly match the funniest captions, followed by positive sentiment. These results are useful for understanding humor and also in the design of more engaging conversational agents in text and multimodal (vision+text) systems. As part of this work, a large set of cartoons and captions is being made available to the community.