The most important research aspect nowadays is the data. Association rule mining is vital mining used in data which mines many eventual informations and associations from enormous databases. Recently researchers focus many research challenges to association rule mining. The first challenge is the generation of the frequent and infrequent itemsets from a large dataset more accurately. Secondly how effectively the positive and negative association rule can be mined from both the frequent and infrequent itemsets with high confidence, good quality, and high comprehensibility with reduced time. Predominantly in existing algorithms the infrequent itemsets is not taken into account or rejected. In recent times it is said that useful information are hidden in this itemsets in the case of medical field. The third challenge are to generate is optimised positive and negative association rule. Several existing algorithms have been implemented in order to assure these challenges but many such algorithms produces data losses, lack of efficiency and accuracy which also results in redundant rules. The major issue in using this analytic optimizing method are specifying the activist initialization limit was the quality of the association rule relays on. The proposed work has three methods which mine an optimized PAR and NAR. The first method is the Apriori_AMLMS (Accurate multi-level minimum support) this algorithm derives the frequent and the infrequent itemsets very accurately based on the user-defined threshold minimum support value. The next method is the GPNAR (Generating Positive and Negative Association Rule) algorithm to mine the PAR and NAR from frequent itemsets and PAR and NAR from infrequent itemsets. The third method are to obtain an optimized PAR and NAR using the decidedly efficient swarm intelligence algorithm called the Advance ABC (Artificial Bee Colony) algorithm which proves that an efficient optimized Positive and negative rule can be mined. The Advance ABC is a Meta heuristic technique stimulated through the natural food foraging behaviour of the honey bee creature. The experimental analysis shows that the proposed algorithm can mine exceedingly high confidence non redundant positive and negative association rule with less time.

I Berin Jeba Jingle1, J Jeya ACelin2
Noorul Islam University, India1, Hindustan College of Arts and Science, India2

Data Mining, Association Rule Mining, Apriori Algorithm, Accurate Multi Level and Multi Support, Advance ABC Algorithm, GPNAR
Published By :
Published In :
ICTACT Journal on Soft Computing
( Volume: 8 , Issue: 2 )
Date of Publication :
January 2018

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.