摘要 |
Disclosed is a method for performing parallel mining of temporal relations in a large capacity event file using a MapReduce model. The method for performing parallel mining of temporal relations in a large capacity event file, according to the present invention, comprises the steps of arranging an event file on the basis of a customer ID and an event time when an event occurs; generating a frequent event type set, satisfying a preset support level or more, from the event file; converting the event file into a frequent event sequence including the frequent event type set; summing up the frequent event sequence to generate a temporal interval file; generating candidate temporal relations in the temporal interval file, and extracting frequent temporal relations satisfying a preset support level or more, from the candidate temporal relations; and generating a temporal relation rule from the frequent temporal relations. |