JokeEval: Are the jokes funny? Review of computational evaluation techniques to improve joke generation

Humor is a complex yet essential aspect of human communication. It can be defined as a communicative expression establishing surprising, incongruent relationships or meanings to amuse. This paper presents empirical evidence demonstrating the successful application of computational methods to humor recognition in AI generated textual data, specifically jokes. Through experiments on synthetic and open-source datasets, we show that automatic classification techniques can effectively differentiate between “Funny” and “Not Funny” jokes. Our results reveal that hybrid Convolutional Neural Networks with Recurrence, trained on high-dimensional vector embeddings of synthetic jokes, achieve a statistically significant F1-Score of 71.2% on the ColBERT dataset. These findings underscore the potential of machine learning approaches in capturing the nuanced nature of humor, paving the way for more sophisticated computational understanding of this fundamental aspect of human interaction and providing a feedback loop for funnier joke generation.