Network Topology Constraint Based Transaction Separation and Mining Algorithm of Telecom Alarm

The efficiency of alarm mining in large scale network is a hot topic in research area of fault location. Many projects have been utilized to improve it and some advance has been achieved, but it is still not very satisfied. Large scale telecom system daily generates large numbers of original alarm records and many of them are redundant information, actually useful information for fault location is only a small part. So in order to increase efficiency of alarm mining, we propose a solution called network topology constraint based transaction separation in this paper and we also compare it with the method of adding topology constraint into mining algorithms. The experiments show that separating transaction with network topology constraint deletes a huge number of redundant alarm information and incorrect transaction sets before executing mining algorithms, therefore, it is can be more efficient and precise than topology constraint based mining algorithm, which filters erroneous frequent sub-patterns during algorithm execution.