Analyzing safety level and recognizing flaws of commercial centers through data mining approach

Abstract

The construction industry, including buildings and commercial centers, is a dynamic industry with diverse and complex nature, which makes its safety provision difficult. The aim of this study is to evaluate the safety status of commercial centers and their classification based on common features; and to uncover the hidden relationships between characteristics of the commercial centers under study by means of data mining techniques. Data required for this study were collected based on a 75-item checklist designed for this study. Indeed, this study included 108 commercial centers. Thereafter, the commercial centers under study were divided into three categories, labeled unsafe, normal, and safe by means of K-means algorithm. The results obtained from the implementation of classification method showed that the two resources, namely, fire protection systems and buildings, played a critical role in the safety of studied commercial centers. The results of in-depth analysis on unsafe commercial centers indicated that these centers have common weaknesses. These weak areas include such items as deficiency of the standards required for the equipment associated with some resources, insufficient training in the use of firefighting equipment, the necessity of the employment of redundant approaches for exit from the building in emergency conditions, and non-feasibility of conducting of operations for firefighting vehicles and lifts. Urban planners and managers and safety officials of the buildings, particularly commercial centers, can apply the results of this study as strategic guidelines.