Predicting Issue Types with seBERT

TL;DR

This paper introduces seBERT, a transformer-based model trained on software engineering data, which outperforms baseline methods in predicting issue types with high accuracy.

Contribution

The paper presents seBERT, a novel pre-trained transformer model specifically trained on software engineering data for issue type prediction, achieving superior performance.

Findings

seBERT achieves an F1-score of 85.7%.

seBERT outperforms fastText baseline in recall and precision.

The model demonstrates strong effectiveness across all issue types.

Abstract

Pre-trained transformer models are the current state-of-the-art for natural language models processing. seBERT is such a model, that was developed based on the BERT architecture, but trained from scratch with software engineering data. We fine-tuned this model for the NLBSE challenge for the task of issue type prediction. Our model dominates the baseline fastText for all three issue types in both recall and precisio} to achieve an overall F1-score of 85.7%, which is an increase of 4.1% over the baseline.