Code Re-use Recommendation Based on Software Requirements Using Natural Language Processing in Python

Code reuse is a pivotal practice in software development, promoting efficiency, reducing redundancy, and ensuring higher quality software products. Recommender systems play a crucial role in this process by suggesting relevant code snippets or components based on specific requirements, thus easing the task of code reuse for developers. In this research paper, we present a Python code re-use recommendation system based on software requirements using Natural Language Processing (NLP). By employing the BERT model, we assess the similarity between software requirements in the e-commerce domain. The resulting software is then pre-processed and evaluated using the CodeBERT model to determine code similarity. Our findings show that similar requirements can indeed produce similar software, promoting efficient code reuse. This system can significantly benefit requirement engineers and developers by enabling the re-utilization of existing code within the same domain, demonstrating the broader applicability of our approach.