Deep security analysis of program code

Abstract Due to the continuous digitalization of our society, distributed and web-based applications become omnipresent making them more secure gains paramount relevance. Deep learning (DL) its representation approach are increasingly been proposed for program code analysis potentially providing a powerful means in software systems less vulnerable. This systematic literature review (SLR) is aiming thorough comparison 32 primary studies on DL-based vulnerability code. We found rich variety approaches, embeddings network topologies. discuss these techniques alternatives detail. By compiling commonalities differences we identify current state research this area future directions. also provide an overview publicly available datasets order foster stronger benchmarking approaches. SLR provides starting point researchers interested deep