Improved association rule mining algorithm for network alarm analysis

With the scale of communication networks becomes increasingly large and the structure becomes increasingly complex, a large number of alarms can be produced every day. As for the fact that the existing algorithms are inefficient and produce too many rules, an improved temporal association rule mining algorithm is proposed in this paper to avoid the dedundent rules and improve the mining efficiency. Simulation experiment with the real dataset from a communication company proves that the proposed algorithm can both reduce the redundant rules and improve the efficiency.