Who's Laughing Now? An Overview of Computational Humour Generation and Explanation

2025-09-25

AI Summary

Computational humor research remains severely underdeveloped, particularly in generating and interpreting non-pun-based humor; mainstream large language models fall far short of human-level performance. Method: This paper systematically surveys state-of-the-art humor generation and comprehension in NLP, emphasizing its stringent demands on commonsense reasoning, contextual modeling, and creative generation. It breaks from the dominant pun-centric paradigm by explicitly incorporating subjectivity and ethical ambiguity as core dimensions, and analyzes modeling bottlenecks using generative AI and LLM techniques. Contribution/Results: The study identifies fundamental limitations in current models—especially regarding contextual sensitivity, implicit intent inference, and cultural adaptability. It clarifies key technical and ethical challenges, and proposes a future research framework anchored on “interpretability, controllability, and ethical alignment,” positioning computational humor as a foundational subfield of NLP.

Technology Category

• Natural Language Processing: Generation
• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Economics, Online Markets and Human Computation: Research challenges in human and human-AI computation
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The creation and perception of humour is a fundamental human trait, positioning its computational understanding as one of the most challenging tasks in natural language processing (NLP). As an abstract, creative, and frequently context-dependent construct, humour requires extensive reasoning to understand and create, making it a pertinent task for assessing the common-sense knowledge and reasoning abilities of modern large language models (LLMs). In this work, we survey the landscape of computational humour as it pertains to the generative tasks of creation and explanation. We observe that, despite the task of understanding humour bearing all the hallmarks of a foundational NLP task, work on generating and explaining humour beyond puns remains sparse, while state-of-the-art models continue to fall short of human capabilities. We bookend our literature survey by motivating the importance of computational humour processing as a subdiscipline of NLP and presenting an extensive discussion of future directions for research in the area that takes into account the subjective and ethically ambiguous nature of humour.

Problem

• Surveying computational humor generation and explanation tasks in NLP
• Assessing LLMs' common-sense knowledge through humor understanding
• Addressing sparse research on non-pun humor generation and explanation

Innovation

• Surveying computational humor generation and explanation techniques
• Evaluating large language models' humor reasoning capabilities
• Identifying research gaps in non-pun humor generation