Comprehensive Evaluation of Vulnerability Mining Tools Based on Grey Relational Analysis

Existing research on the capability evaluation of vulnerability discovery tools predominantly focuses on case-specific metric comparisons and analyses, lacking generalizable evaluation methodologies to provide holistic assessment conclusions for guiding tool selection. This paper investigates metric systems and evaluation methodologies for the comprehensive assessment of vulnerability discovery tools. First, we systematically reviewed 58 relevant studies from the past decade to establish a robust metrics set. Next, we propose a hybrid weighting approach integrating subjective and objective prioritization mechanisms to assign metric weights. Finally, a grey relational analysis algorithm was applied to synthesize test data for multidimensional evaluation. Experimental results demonstrate the effectiveness of the proposed combined weighting method and grey relational analysis in enabling systematic, comprehensive assessments of vulnerability discovery tools.