ColBERT: Using BERT Sentence Embedding for Humor Detection

27 Apr 2020  ·  Issa Annamoradnejad, Gohar Zoghi ·

Automatic humor detection has interesting use cases in modern technologies, such as chatbots and virtual assistants. In this paper, we propose a novel approach for detecting humor in short texts based on the general linguistic structure of humor. Our proposed method uses BERT to generate embeddings for sentences of a given text and uses these embeddings as inputs of parallel lines of hidden layers in a neural network. These lines are finally concatenated to predict the target value. For evaluation purposes, we created a new dataset for humor detection consisting of 200k formal short texts (100k positive and 100k negative). Experimental results show that our proposed method can determine humor in short texts with accuracy and an F1-score of 98.2 percent. Our 8-layer model with 110M parameters outperforms the baseline models with a large margin, showing the importance of utilizing linguistic structure of texts in machine learning models.

PDF Abstract

Datasets


  Add Datasets introduced or used in this paper
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Humor Detection 200k Short Texts for Humor Detection ColBERT model F1-score 0.982 # 1
Humor Detection 200k Short Texts for Humor Detection Decision Tree F1-score 0.794 # 6
Humor Detection 200k Short Texts for Humor Detection SVM F1-score 0.874 # 4
Humor Detection 200k Short Texts for Humor Detection Multinomial NB F1-score 0.882 # 3

Methods