computational-humor

Deep-learning model presented in "DataStories at SemEval-2017 Task 6: Siamese LSTM with Attention for Humorous Text Comparison".

Reproducible builder for the Oogiri-Master benchmark and Oogiri-Corpus humor datasets, sourced from Japanese Oogiri pages.

Implementation of the paper "How Did This Get Funded?! Automatically Identifying Quirky Scientific Achievements"

Automatic generation of a popular Thai wordplay joke "May I ... you, even as a ...?"

DS@GT ARC's CLEF 2025 JOKER Task 2 submission. Sentence-transformer + FAISS retrieval, LLM-generated contrastive non-pun rewrites, and an LLM generator/discriminator pair for English→French pun translation.

A Tool for Analyzing Short Stories Using the General Theory of Verbal Humor

English to French pun translation system. FAISS-indexed dense retrieval over a 370K-entry French expression bank and a BGE-M3 embedding model fine-tuned on 2.7M IPA phonetic relations, feeding Claude, Gemini, and GPT as competing pun generators, ranked by weighted Borda voting across four LLM-judge personas.