JokeMeter at SemEval-2020 Task 7: Convolutional humor

TL;DR

This paper presents JokeMeter, a convolutional neural network system developed for humor evaluation in SemEval-2020 Task 7, analyzing learned features to understand humor detection.

Contribution

The paper introduces a CNN-based approach for humor detection and provides insights into the learned features of the model.

Findings

Effective humor evaluation with CNN architecture

Analysis of internal features learned by the model

Insights into humor detection mechanisms

Abstract

This paper describes our system that was designed for Humor evaluation within the SemEval-2020 Task 7. The system is based on convolutional neural network architecture. We investigate the system on the official dataset, and we provide more insight to model itself to see how the learned inner features look.