Transformer based Neural Joke Generator
Taaha Kazi, Sameer Joshi, Steeve Kaitharath, Imran Ali Mirza · International Journal of Computer Applications · 2021
Humor is a complex and intrinsic part of human conversation, which involves a deep understanding of grammatical structure and knowledge of the world.Building computational models that can identify and generate humor remains a challenging field.This work presents a neural network based joke generator that employs a transformer-based architecture.To improve the generator's performance, the model was further trained with Proximal Policy Optimization (PPO), a reinforcement learning algorithm.The model's performance was evaluated by human ratings by conductingqualitative analysis.