Information about the factors that lead to the deterioration of bridges is essential for bridge maintenance. Pinpointing what these factors are will certainly enhance the effectiveness of bridge management. However, a review of the literature reveals that such factors are mainly determined based on experts' opinions rather than a systematic approach. In this study the factors leading to deterioration of RC bridge decks are grouped into six common types. Twenty-nine candidate factors are selected from an extensive review of past work as well as from the inventory of the Taiwan Bridge Management System. A data mining technique, the Rough Set Theory (RST), is employed to find the factors that have the most significant impact on deterioration. It is found that weather-related factors are rather significant for almost all types of deterioration. Finally, the factors mined by RST are compared to those obtained by Mann-Whitney U (MWU). The results of comparison appear fairly consistent, which validates the proposed approach.
COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING