Neuro-symbolic Zero-Shot Code Cloning with Cross-Language Intermediate Representation
This addresses the challenge of code cloning for legacy systems like COBOL, which is incremental as it adapts existing methods to a new domain.
The paper tackles the problem of finding semantically similar code clones for the legacy language COBOL without training data by using a neuro-symbolic approach with a cross-language intermediate representation, achieving a 12.85 MAP@2 improvement over a pre-trained model in a zero-shot setting.
In this paper, we define a neuro-symbolic approach to address the task of finding semantically similar clones for the codes of the legacy programming language COBOL, without training data. We define a meta-model that is instantiated to have an Intermediate Representation (IR) in the form of Abstract Syntax Trees (ASTs) common across codes in C and COBOL. We linearize the IRs using Structure Based Traversal (SBT) to create sequential inputs. We further fine-tune UnixCoder, the best-performing model for zero-shot cross-programming language code search, for the Code Cloning task with the SBT IRs of C code-pairs, available in the CodeNet dataset. This allows us to learn latent representations for the IRs of the C codes, which are transferable to the IRs of the COBOL codes. With this fine-tuned UnixCoder, we get a performance improvement of 12.85 MAP@2 over the pre-trained UniXCoder model, in a zero-shot setting, on the COBOL test split synthesized from the CodeNet dataset. This demonstrates the efficacy of our meta-model based approach to facilitate cross-programming language transfer.