An empirical study of developer-ChatGPT refactoring chats yields a 25-theme taxonomy, apology/affirmation signals, and a structured prompt template that reduces conversation turns.
On the relation of refactorings and software defect prediction,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
ChatGPT for Code Refactoring: Analyzing Topics, Interaction, and Effective Prompts
An empirical study of developer-ChatGPT refactoring chats yields a 25-theme taxonomy, apology/affirmation signals, and a structured prompt template that reduces conversation turns.