An overview of interval encoded temporal mining involving prioritized mining, fuzzy mining, and positive and negative rule mining

ISSN:0975-9646 Databases and data warehouses have become a vital part of many organizations. So useful information and helpful knowledge have to be mined from transactions. In real life, media information has time attributes either implicitly or explicitly called as temporal data. This paper focuses on an encoding method for the temporal database that reduces the memory utilization during processing. The first approach involves temporal mining applying the conventional algorithms like Apriori, AprioriTid and AprioriHybrid to an encoded temporal database that has a better performance than that when applied over a static database. The second approach involves weighted temporal mining over an encoded temporal database consisting of items which are prioritized by assigning weights. These weights are given according to the importance of the item from the user’s perspective. A fuzzy mining approach involving AprioriTid for weighted association rule mining gives better results than quantitative values. Also a method for positive and negative temporal mining extends traditional associations to include association rules of forms A=> ¬B, ¬A => ¬B, A=> ¬B, which indicate negative associations between itemsets. The experimental results are drawn from the complaints database of the telecommunication system which presents the most feasible temporal mining method with reduced time and computational complexities.