arXiv 2408.04663
Dopamin: Transformer-based Comment Classifiers through Domain Post-Training and Multi-level Layer Aggregation
By Nam Le Hai and Nghi D. Q. Bui
Published 2024-08-06
Citation lineage
Review the prior work and downstream research connected to this paper.
Code comments provide important information for understanding the source code. They can help developers understand the overall purpose of a function or class, as well as identify bugs and technical debt. However, an overabundance of comments is meaningless and counterproductive. As a result, it is critical to automatically filter out these comments for specific purposes. In this paper, we present Dopamin, a Transfor…