MASTERÂ Assignment
Process optimization of developing vulnerability datasets through systematic data consolidation.
Type : Master M-CS
Period: February - September, 2026
Student: Waghmare S. (Soham, Student M-CS)
Date Final project: September 14, 2026
Supervisors:
Abstract:
Hospital and other resource-constrained organizations rely on the National Vulnerability Database (NVD) to prioritize security remediation, but NVD alone lacks exploitability context, weakness classification depth, and known-exploitation status, making it difficult for security teams to make timely and informed decisions. This study addresses this problem by developing a framework that contains an automated pipeline for integrating vulnerability data from multiple public sources into a set of structured datasets. Guided by the Design Science Research Methodology (DSRM) and implemented using CRISP-DM, data from five public sources - NVD, CISA KEV, FIRST.org EPSS, MITRE CWE, and CVE.org, were extracted, aggregated, and enriched into eight structured datasets and imported into a queryable Neo4j graph database. This artifact was evaluated in two stages: a technical verification of its completeness and linkability, and a scenario-based evaluation using the Technology Acceptance Model (TAM), for which participants assessed the artifact against three realistic use cases. The result of this was a functioning framework demonstrating strong completeness across relationship types and consistently positive TAM scores, with case comparisons showing substantial gaps against an NVD-only baseline. This framework can be used by organizations to extend vulnerability coverage beyond NVD and to support faster, more informed vulnerability prioritization.