Regresar

Techssn at haha @ iberlef 2021: Humor detection and funniness score pbkp_rediction using deep learning techniques

Abstract:

This paper is a description of a system used to classify tweets in Spanish as humorous or not and rate the level of humor of each tweet. The system developed by the team TECHSSN uses binary classification techniques to classify the text as humor or not (subtask1) and ensemble learning regression model to rate the funniness score of the tweet (subtask2). The data undergoes preprocessing and is given to a modification of BERT [1] (Bidirectional Encoder Representations from Transformers) for the subtask1. The model is retrained, and the weights are learned for the dataset provided. XGBoost ensemble model is used to pbkp_redict the funniness score on the BERT output for subtask 2. These systems were developed for the HAHA subtasks for IberLEF2021.