Alarm association rules mining based on run log for civil aviation information system
Wang Jing, Huaichao Wang, Jiyang Zhang · 2017
The new generation of Civil Aviation Passenger Service Information System is a national key information system, its stable and reliable operation must be ensured. For the great challenge of safety monitoring caused by the complexity of its function and application, an alarm association rules mining scheme for complex information system is proposed in this paper, and the algorithm of alarm association rules mining based on double constrained sliding time window model is studied emphatically. Firstly, the characteristics of the alarm data in monitoring log of Travelsky Passenger Information and Service system is analyzed to structure these data. Then, the sliding time window model based on the double constraint is introduced to divide the alarm events in the alarm database, and a matrix Apriori algorithm is used to mine the alarm association rules by using the divided log alarm event data, so as to obtain the anomalous event alarm association between complex applications. Finally, the experimental results show that the method has more efficient and reasonable mining results.