HumorPlanSearch: Structured Planning and HuCoT for Contextual AI Humor

TL;DR

HumorPlanSearch is a modular AI humor generation pipeline that explicitly models context, cultural background, and stylistic reasoning to produce more coherent, adaptive, and culturally sensitive jokes, outperforming baseline methods.

Contribution

It introduces a novel structured pipeline combining plan search, humor reasoning templates, knowledge graphs, and iterative revision to enhance contextual humor generation.

Findings

Full pipeline improves Humor Generation Score by 15.4%

Context-aware methods outperform baseline humor models

Human judges rate humor as more coherent and culturally attuned

Abstract

Automated humor generation with Large Language Models (LLMs) often yields jokes that feel generic, repetitive, or tone-deaf because humor is deeply situated and hinges on the listener's cultural background, mindset, and immediate context. We introduce HumorPlanSearch, a modular pipeline that explicitly models context through: (1) Plan-Search for diverse, topic-tailored strategies; (2) Humor Chain-of-Thought (HuCoT) templates capturing cultural and stylistic reasoning; (3) a Knowledge Graph to retrieve and adapt high-performing historical strategies; (4) novelty filtering via semantic embeddings; and (5) an iterative judge-driven revision loop. To evaluate context sensitivity and comedic quality, we propose the Humor Generation Score (HGS), which fuses direct ratings, multi-persona feedback, pairwise win-rates, and topic relevance. In experiments across nine topics with feedback from 13 human judges, our full pipeline (KG + Revision) boosts mean HGS by 15.4 percent (p < 0.05) over a strong baseline. By foregrounding context at every stage from strategy planning to multi-signal evaluation, HumorPlanSearch advances AI-driven humor toward more coherent, adaptive, and culturally attuned comedy.