International Journal of Engineering and Technology

The field of data mining has flourished into research area of significant technological and social importance due to the advancement in technology. Mining frequent pattern or itemset from real data environment is a fundamental and essential problem in many data mining applications. The Apriori-inspired algorithms show good performance with sparse datasets such as market-basket data, where the frequent patterns are very short. However, in the area with dense datasets such as telecommunication, computational biology and census data, the performance of these algorithms degrades incredibly as there were many, long frequent patterns. The focus of this paper is to design a CPU- efficient algorithm CIMTEL for finding closed frequent calling patterns (long pattern) in a telecommunication database. Due to the evaluation of next generation telecommunication network the amount of communication data’s rises in volume, vaiety and velocity (Bid Data). Thus algorithm provides an way for analysis the consumer for service provider. The Performance of this algorithm outperforms the former COLTEL , CHARM and EXPEDITE algorithm by an order of two for a worst case scenario. Also the performance analysis of this algorithm with former algorithms is determined.