Computational Humor

1 minute read

Published:

Cycle of Defining Humor

Introduction

“Humor is instant vacation” - Milton Berle

  1. What is humor?
  2. what are the important computational theories of humor?
  3. What are the popular models and their current limitations?
  4. What are the popular benchmarks?
  5. How do we accurately measure the performance of such systems?
  6. What are the novel applications enabled by these systems?

*(Under Construction)

Datasets for Humor Detection

Dataset NameData TypeDescription and Characteristics
CMU Multimodal Opinion Sentiment and Emotion (MOSEI)Audio, Visual, TextA large-scale dataset containing audio, visual, and textual data for sentiment, emotion, and humor analysis. It includes speech, text transcriptions, and visual features from movies.
YouTube Multimodal Humor Detection DatasetAudio, Video, TextA multimodal dataset that combines audio, video, and text information from YouTube videos with humorous content.
MovieHumorDetection (MHD) DatasetAudio, VideoThis dataset focuses on humor detection in movie scenes, providing audio and visual cues from various films.
Reddit Jokes DatasetTextA collection of humorous text data from Reddit. Used for text-based humor detection research.

References

  1. https://en.wikipedia.org/wiki/Theories_of_humor
  2. https://floriandietz.me/humor/
  3. Computers Learning Humor is No Joke
  4. A Little Metatheory: thoughts on what a theory of computational humor should look like
  5. Mulit modal humor detection
  6. Reddit Humor Detection
  7. UR-Funny