Embodiments of the invention are directed to systems, methods, and computer program products for a security vulnerability analysis and management platform utilizing deep learning technology and knowledge graph database structures. The system is configured to receive software code and metadata corresponding to existing issues and defects present in the software code or associated with the implementation of the software code. By applying a deep learning technique to extract data from the open source software code, wherein the data corresponds to potential issues and defects in the open source software code, the system is configured to populate a knowledge graph database and build a unified cybersecurity ontology that can later be organized and queried based on user input.