MagicPai at SemEval-2021 Task 7: Method for Detecting and Rating Humor Based on Multi-Task Adversarial Training

TL;DR

This paper presents MagicPai's multi-task adversarial training approach for SemEval 2021 Task 7, effectively detecting and rating humor and offense in text by enhancing robustness and capturing task relationships.

Contribution

Introduces a multi-task learning model with adversarial examples and confidence-based loss correction for humor detection and rating.

Findings

Effective in humor detection and rating tasks

Improves robustness through adversarial perturbations

Captures relationships between humor presence and intensity

Abstract

This paper describes MagicPai's system for SemEval 2021 Task 7, HaHackathon: Detecting and Rating Humor and Offense. This task aims to detect whether the text is humorous and how humorous it is. There are four subtasks in the competition. In this paper, we mainly present our solution, a multi-task learning model based on adversarial examples, for task 1a and 1b. More specifically, we first vectorize the cleaned dataset and add the perturbation to obtain more robust embedding representations. We then correct the loss via the confidence level. Finally, we perform interactive joint learning on multiple tasks to capture the relationship between whether the text is humorous and how humorous it is. The final result shows the effectiveness of our system.