Deep Vision-Data Mining To Find Insights and Visualization in Code Repositories

Abstract

Deep Vision is a code mining system for analyzing and visualizing a repository's codebase so that its users may obtain a sense of the repository's insights. This system will examine codebases and support as many languages as feasible. This system visualizes the file structure, vocabulary and length change rates, comprehensibility and defect rates, etc. It is vital to have a comprehensive grasp of the codebase to manage the program's complexity by calculating multiple factors and presenting them in a descriptive and engaging dashboard to enhance the quality of the software process and the project's controllability. Improved code visualization may help improve code understandability while lowering development costs. In addition, our visualization regions and methodologies are one-of-a-kinds. To get rapid and reliable results, we will create new machine learning models and algorithms for analysis and new categories of a code repository. Our dataset for this research will be GitHub open-source code repositories