Abstract:

Currently those algorithms to mine the alarm association rules are limited to the minimal support, so that they can only obtain the association rules among the frequently occurring alarm events, To address this problem, a new mining algorithm based on the statistical correlation was proposed, which firstly acquired the alarm net units with the same character by clustering; and then discovered the association rules from both high-frequency and low-frequency alarm events with the high correlativity and the high confidence. Experimental results demonstrated that this algorithm was efficient and accurate to mine the association rules among alarm events with both high-frequency and low-frequency.

Key words: fault management, association rules, data mining, correlation