CodeBLEU: a Method for Automatic Evaluation of Code Synthesis

TL;DR

This paper introduces CodeBLEU, a new automatic evaluation metric for code synthesis that incorporates syntax and semantics to better correlate with human judgment than traditional metrics like BLEU.

Contribution

The paper proposes CodeBLEU, which enhances BLEU with syntax and semantic features, providing a more accurate automatic evaluation for code generation tasks.

Findings

CodeBLEU correlates better with human judgment than BLEU.

It improves evaluation accuracy across multiple code synthesis tasks.

Experimental results validate the effectiveness of CodeBLEU.

Abstract

Evaluation metrics play a vital role in the growth of an area as it defines the standard of distinguishing between good and bad models. In the area of code synthesis, the commonly used evaluation metric is BLEU or perfect accuracy, but they are not suitable enough to evaluate codes, because BLEU is originally designed to evaluate the natural language, neglecting important syntactic and semantic features of codes, and perfect accuracy is too strict thus it underestimates different outputs with the same semantic logic. To remedy this, we introduce a new automatic evaluation metric, dubbed CodeBLEU. It absorbs the strength of BLEU in the n-gram match and further injects code syntax via abstract syntax trees (AST) and code semantics via data-flow. We conduct experiments by evaluating the correlation coefficient between CodeBLEU and quality scores assigned by the programmers on three code synthesis tasks, i.e., text-to-code, code translation, and code refinement. Experimental results show that our proposed CodeBLEU can achieve a better correlation with programmer assigned scores compared with BLEU and accuracy.