ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 7. Vol. 28. 2022

DOI: 10.17587/it.28.378-391

S. M. Avdoshin, Ph. D., Professor, G. A. Arutyunov, Master's Student,  HSE University, Moscow, Russian Federation

Code Analysis and Generation Methods using Neural Networks: An Overview

The global pandemic has outlined the shortfall of human resources in the information technology sector. On the estimation of analysts, the labor shortage of IT-specialists in Russia in 2021 is between 500 thousand and 1 million people. Educating and bringing to market such numerous personnel may take years. The task of optimizing the process of the development of IT-projects is as urgent as ever. This includes automizing repetitive tasks performed by developers. This article presents an overview of the methods of using neural networks in the field of generation and analysis program source code.
Keywords: neural network technologies, artificial intelligence, graph neural networks, code analysis, code generation methods, machine programming, reinforcement learning

P. 378–391

To the contents