Semi-supervised Learning for Source Code Function Classification Using Hierarchical Density-Based Clustering

In the field of software development, source code topic classification is critical for program comprehension and software maintenance. Source code topic classification is a challenging task that involves classifying software artifacts into predefined topics. Traditional supervised learning approaches require a large amount of labeled data, which is often expensive and time-consuming to obtain. In this paper, we explore the application of semi-supervised learning techniques for source code and propose a semi-supervised learning approach that leverages both labeled and unlabeled data to improve the accuracy of source code topic classification. We utilize hierarchical density-based clustering to group similar source code artifacts, and then apply semi-supervised learning techniques to assign topics to these code functions. Experimental results on a real-world dataset demonstrate the effectiveness of our proposed approach in achieving high accuracy and less complexity, which is due to using semi-supervised learning.