Abstract:
In this work, we performed a seismic vulnerability assessment of the city of Constantine (Algeria)
using A classic European macro-seismic method Risk-UE and innovative data mining-based methods
(Association Rule Learning, ARL).
The ARL method consists in establishing relationships between building attributes (number of stories
or building age) and the vulnerability classes of the European Macroseismic Scale, EMS 98. This
approach avoids the costly process in the analysis of seismic vulnerability requires site surveys for
collecting the necessary building characteristics, which often discourages the assessment of seismic
risk initiatives in moderate seismic prone regions. We validate our learning (proxy) by comparing
estimated vulnerability classes by our simple method to those actually observed on site by the
classical method Risk-UE.
For a given seismic scenario, the results give the likely damage comparable to those obtained by the
traditional method. These results are presented as histograms or risk maps using geographic
information system (GIS).
Finally, these results can be used by many policymakers as insurers, planners and urban
development, responsible for public safety and seismic specialists working on the codes of protection.
