Literature survey of deep learning-based vulnerability analysis on source code

Abstract

Vulnerabilities in software source code are one of the critical issues in the realm of software code auditing. Due to their high impact, several approaches have been studied in the past few years to mitigate the damages from such vulnerabilities. Among the approaches, deep learning has gained popularity throughout the years to address such issues. In this literature survey, the authors provide an extensive review of the many works in the field software vulnerability analysis that utilise deep learning‐based techniques. The reviewed works are systemised according to their objectives (i.e. the type of vulnerability analysis aspect), the area of focus (i.e. the focus area of the analysis), what information about source code is used (i.e. the features), and what deep learning techniques they employ (i.e. what algorithm is used to process the input and produce the output). They also study the limitations of the papers and topical trends concerning vulnerability analysis.

Question and Answer Techniques for Financial Audits in Universities Based on Deep Learning

“…In recent years, China's nancial auditing of colleges and universities has been developed in an unprecedented way by combining national conditions, absorbing international advanced ideas, and groping on the road of practice, constantly improving and re ning. With the ourishing development of computer science and technology, audit informatization has become a wave that promotes the development and progress in the eld of nancial auditing in universities, improving audit e ciency, and saving audit costs [1].…”

Software vulnerability prediction: A systematic mapping study

“…In particular, they reviewed 61 primary studies, published up to February 2021, and explored the data requirements, collection, labeling and cleaning processes, also discussing the challenges of these processes. Semasaba et al presented a systematic literature review of software vulnerability analysis on source code using Deep Learning (DL) [5]. They retrieved 28 papers published between 2014 and 2020 by searching on six widely used Digital Libraries for studies about software vulnerability detection.…”

Data Preparation for Software Vulnerability Prediction: A Systematic Literature Review

“…To the best of our knowledge, only two studies have been published that focus on systematically reviewing SVP research and knowledge: Semasaba et al [17], and Hasif et al [4]. The former exclusively investigated Deep Learning techniques, whereas the latter provided a wider view of SV detection, including non-learning based techniques.…”

“…The SV data used to train a model is the most imperative component of this data-driven process. Although most studies have reported data preparation and data quality as significant issues for this research area [4], [10], [13], [16], [17], [24], they have not performed in-depth analysis of the data quality in SVP research to determine the encountered issues or potential solutions. This knowledge gap fails to provide practitioners and researchers with the specific insights needed to remediate data quality issues.…”

“…However, despite the importance and difficulties of SV data preparation for both industry and academia, there has been relatively little effort allocated to systematically understand the known challenges of data preparation for SVP models and how to address them. Whilst there are several secondary studies that have analyzed SVP research [4], [13], [16], [17] and acknowledged the existence of problems with the data preparation, these studies have not focused on thoroughly investigating the data preparation related challenges in SVP research.…”