In the ever evolving field of software development, understanding and maintaining complex codebases is crucial. There exist machine learning models and algorithms that aid in this by specifically learning to ‘understand code’, allowing engineers to build applications that help develop and maintain these large codebases. Although existing source-code machine learning models often overlook an important factor: the code's context. Our research focuses on leveraging external contextual information to enhance source-code model performance. We’ve developed a data pipeline that utilizes CodeQL to extract contextual information from the CodeSearchNet benchmark dataset to extend and create an augmented version of the dataset. We also experiment with a model architecture, CodeContext, that integrates context with code snippets in an effective manner. This approach promises to enhance code comprehension and maintenance, marking a significant advancement in software development tools.
↧