The bridge has been a crucial element of the transportation system of the U.S.A. for many years. The National Bridge Inventory (NBI) reported more than 615,000 national bridges in 2018. Maintaining and fixing bridges is a crucial task for transportation agencies to keep the road network connected. Louisiana, which has 12,899 bridges, was selected as the study site for this study. The American Road and Transportation Builders Association (ARTBA) reported in 2019 that 13% of all Louisiana bridges were classified as structurally deficient. This study applies a data mining algorithm, the empirical Bayes geometric mean (EBGM) method, to identify critical patterns of the bridge inventory condition at element level as a measure of vulnerability, using NBI rating data from 2015 to 2018. It finds that severe condition is highly associated with the following elements, regardless of their structural importance: bridge joints, and bridge rail timber,bearing other, and superstructure floor beam reinforced concrete elements. Poor condition is highly associated with elements like top flange reinforced concrete,bearing movable, and superstructure floor beam reinforced concrete. The quantification scores developed in this study could help transportation agencies and bridge engineers to identify more easily the key element or combination of elements associated with poor or severe condition, so that they can make data-driven decisions in maintaining and repairing the most needed bridge elements.