Witscript: A System for Generating Improvised Jokes in a Conversation

TL;DR

Witscript is a novel joke generation system that creates contextually relevant, improvised jokes for chatbots by combining joke writing algorithms, NLP keyword extraction, wordplay, and neural network completion, improving chatbot humor.

Contribution

The paper introduces Witscript, a new system integrating joke writing algorithms with neural networks for real-time, context-aware joke improvisation in chatbots, a novel approach in chatbot humor generation.

Findings

Human evaluators judged over 40% of responses as jokes.

Witscript successfully generates contextually relevant jokes in conversations.

The system advances chatbot humor capabilities with a combination of NLP and neural models.

Abstract

A chatbot is perceived as more humanlike and likeable if it includes some jokes in its output. But most existing joke generators were not designed to be integrated into chatbots. This paper presents Witscript, a novel joke generation system that can improvise original, contextually relevant jokes, such as humorous responses during a conversation. The system is based on joke writing algorithms created by an expert comedy writer. Witscript employs well-known tools of natural language processing to extract keywords from a topic sentence and, using wordplay, to link those keywords and related words to create a punch line. Then a pretrained neural network language model that has been fine-tuned on a dataset of TV show monologue jokes is used to complete the joke response by filling the gap between the topic sentence and the punch line. A method of internal scoring filters out jokes that don't meet a preset standard of quality. Human evaluators judged Witscript's responses to input sentences to be jokes more than 40% of the time. This is evidence that Witscript represents an important next step toward giving a chatbot a humanlike sense of humor.