Analysis of deep learning in software engineering

Deep learning has been increasingly popular in the area of information engineering over recent years (SE). However, there are so many unfinished concerns to be studied. How can researchers incorporate deep learning into SE issues? What SE phases are supported by in-depth learning? Are the professionals benefiting from deep learning? Responses allow clinicians and researchers to build functional deep learning models for SE activities. To address these questions, we perform a bibliography review of 98 SE academic articles using deep learning techniques. We notice that 41 SE activities have been enabled by deep learning optimised solutions in all SE phases. In which 84.7 percent of papers only use traditional deep learning techniques and their variants to overcome Software Engineering problems. Practicability is becoming a problem with the use of deep learning methods. More Software Engineering researchers may be attracted in the future to Enhance the effectiveness, efficiency, comprehensibility and testability with deep learning- based methods.