New algorithms for fast discovery of association rules (original) (raw)

KSD an Efficient Algorithm for Association Rule Mining

Abstract — Data mining is an emerging field that comprises of various functions like classification, association rule mining, clustering, and outlier analysis. Association rule mining is a major, interesting and extremely studied function of data mining. Association rule mining identifies the correlation between different itemsets and find frequent and interesting rules. Frequent itemset mining is very common first step in considering datasets through wide range of applications. There have been proposed some methods in literature which scan database twice or more times to find approximate frequent patterns and frequent itemsets. Scanning database again and again makes mining process tedious and slow. The traditional approaches needs that every item in itemset happens in each supporting transaction. Yet the actual data has noise (meaningless data) and in existence of a noise, outdated itemset mining procedures might not be able to identify related frequent itemset(s). We have proposed a method in this paper that solved above mentioned problems. It scans database only once and makes mining fast and efficient. Our proposed method used technique named Fault Tolerance to handle noisy data and replaced database with a tree like structure. We are unaware of any technique yet introduced that can find approximate frequent itemset with only one scan of database. Further, our proposed method has an advantage on traditional Apriori and frequent pattern (FP) Tree) method as for as scanning and infrequent candidate generation are concerned. Keywords: Approximate pattern; frequent pattern; Apriori; fault tolerance; FP-Tree; FT-Apriori; AFI-FP

Discovering Frequent Closed Itemsets for Association Rules

1999

In this paper, we address the problem of finding frequent itemsets in a database. Using the closed itemset lattice framework, we show that this problem can be reduced to the problem of finding frequent closed itemsets. Based on this statement, we can construct efficient data mining algorithms by limiting the search space to the closed itemset lattice rather than the subset lattice. Moreover, we show that the set of all frequent closed itemsets suffices to determine a reduced set of association rules, thus addressing another important data mining problem: limiting the number of rules produced without information loss.We propose a new algorithm, called A-Close, using a closure mechanism to find frequent closed itemsets. We realized experiments to compare our approach to the commonly used frequent itemset search approach. Those experiments showed that our approach is very valuable for dense and/or correlated data that represent an important part of existing databases.

Survey on Incremental Association Rule Mining to Find Frequent Itemsets

2015

Mining frequent Itemsets has proved to be very difficult because of its computational complexity. But, , it has gained a lot of popularity due to the usefulness of association rules, despite having huge processing cost. This paper provides a comprehensive survey on the state-of-art algorithms for association rule mining, specially when the datasets used for rule mining are dynamic. When new data are added to a original dataset it may lead to additional rules or to modification of some existing rules. To find the association rules from the whole (old as well as new) dataset will be wastage of time only if the process is restarted from the beginning. Several algorithms have been developed to attend this important issue of the association rule mining problem. This paper analyzes some of the algorithms to tackle the incremental association rule mining problem.

A Survey on Association Rule Mining Algorithms for Frequent Itemsets

— These days many current data mining tasks are accomplished successfully only in discovery of Association rule. It appeals more attention in frequent pattern mining because of its wide applicability. Many researchers successfully presented several efficient algorithms with its performances in the area of rule generation. This paper mainly assembles a theoretical survey of the existing algorithms. Here author provides the considered Association rule mining algorithms by beginning an overview of some of the latest research works done on this area. Finally, discusses and concludes the merits and limitation.

Mining Association Rules from Infrequent Itemsets: A Survey

International Journal of Innovative Research in Science, Engineering and Technology, 2013

Association Rule Mining (AM) is one of the most popular data mining techniques. Association rule mining generates a large number of rules based on support and confidence. However, post analysis is required to obtain interesting rules as many of the generated rules are useless.However, the size of the database can be very large. It is very time consuming to find all the association rules from a large database, and users may be only interested in the associations among some items.So mining association rules in such a way that we maximize the occurrences of useful pattern. In this paper we study several aspects in this direction and analyze the previous research.So that we come with the advantages and disadvantages.