期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：3
页码：4485
DOI：10.15680/IJIRCCE.2017.0503161
出版社：S&S Publications
摘要：Data mining is the procedure of extracting patterns and knowledge from large dataset. Using differentmembership functions for different levels, the quantitative data is converted into fuzzy values. In quantitative datasetinteresting rules are mined using minimum support and minimum confidence. These threshold values are difficult to set foreach dataset. To overcome the above problem, the fuzzy temporal coherent rule mining algorithm has been introduced for apredefined taxonomy. The association in the same level items isfound using coherent logic. Fuzzy value is generated foreach level using different low, middle and high membership values (Fig. 1). Positive and negative coherencies are foundusing coherency formula. Coherency is found between level-wise items which include items at the same level and crosslevel. In this methodology, both positive and negative coherent rules are mined for same level in a taxonomical dataset.From the resultant data, the maximum number of coherent rules for the items among the levels in different time periodsisalso found. The rules mined gives knowledge on association of items without the usage of interesting measure.
