Against-Expectation Pattern Discovery: Identifying Interactions within Items with Large Relative-Contrasts in Databases

Abstract — We design a new algorithm for identifying against- expectation patterns. An against-expectation pattern is either an itemset whose support is out of a range of the expected support value, referred to as an against-expectation itemset, or it is an association rule generated by an against-expectation itemset, referred to as an against-expectation rule. Therefore, against- expectation patterns are interactions within those items whose supports have large relative-contrasts in a given database. We evaluate our algorithms experimentally, and demonstrate that our approach is efficient and promising Index Terms — Exception, against-expectation pattern, nearest- neighbor graph, correlation analysis. I. I NTRODUCTION RADITIONALLY , association analysis has focused on techniques aimed at discovering interactions within data. It has mainly involved association rules [1,4,21] and negative association rules [18,23]. These rules can be identified from data by using statistical methods and grouping. In real world applications, data marketers seek to identify interactions and predict profit potential in the relative-contrast of sales. Meanwhile, they recognise that principle items, having large relative-contrasts with respect to their supports expected for a given database, may provide larger profit potential than those with low relative-contrasts. In this paper, we refer to interactions within items that have large relative-contrast as against-expectation patterns. Up until now, the techniques for mining against-expectation patterns have been undeveloped. To rectify this, our paper studies the issue of mining against-expectation patterns in databases. An against-expectation pattern is either an itemset whose support is out of a range of the expected support value (expectation), referred to here as an against-expectation itemset, or an association rule generated from against-expectation itemsets, referred to as an against-expectation rule.

[1]  Leo Egghe,et al.  Construction of weak and strong similarity measures for ordered sets of documents using fuzzy set techniques , 2003, Inf. Process. Manag..

[2]  Jeffrey Xu Yu,et al.  Mining Changes of Classification by Correspondence Tracing , 2003, SDM.

[3]  Xindong Wu,et al.  Mining Both Positive and Negative Association Rules , 2002, ICML.

[4]  Rajeev Motwani,et al.  Beyond market baskets: generalizing association rules to correlations , 1997, SIGMOD '97.

[5]  Vipin Kumar,et al.  Chameleon: Hierarchical Clustering Using Dynamic Modeling , 1999, Computer.

[6]  Yoon Ho Cho,et al.  Mining changes in customer buying behavior for collaborative recommendations , 2005, Expert Syst. Appl..

[7]  Hongjun Lu,et al.  Exception Rule Mining with a Relative Interestingness Measure , 2000, PAKDD.

[8]  Balaji Padmanabhan,et al.  A Belief-Driven Method for Discovering Unexpected Patterns , 1998, KDD.

[9]  Geoffrey I. Webb.  Efficient search for association rules , 2000, KDD '00.

[10]  Jian Tang,et al.  Mining exception instances to facilitate workflow exception handling , 1999, Proceedings. 6th International Conference on Advanced Systems for Advanced Applications.

[11]  Tomasz Imielinski,et al.  Mining association rules between sets of items in large databases , 1993, SIGMOD Conference.

[12]  Gregory Piatetsky-Shapiro,et al.  Discovery, Analysis, and Presentation of Strong Rules , 1991, Knowledge Discovery in Databases.

[13]  Hongjun Lu,et al.  Efficient Search of Reliable Exceptions , 1999, PAKDD.

[14]  George Karypis,et al.  C HAMELEON : A Hierarchical Clustering Algorithm Using Dynamic Modeling , 1999 .

[15]  Shamkant B. Navathe,et al.  Mining for strong negative associations in a large database of customer transactions , 1998, Proceedings 14th International Conference on Data Engineering.

[16]  Wynne Hsu,et al.  Mining Changes for Real-Life Applications , 2000, DaWaK.

[17]  Howard J. Hamilton,et al.  Extracting Share Frequent Itemsets with Infrequent Subsets , 2003, Data Mining and Knowledge Discovery.

[18]  Stephen D. Bay,et al.  Detecting Group Differences: Mining Contrast Sets , 2001, Data Mining and Knowledge Discovery.

[19]  Wynne Hsu,et al.  Finding Interesting Patterns Using User Expectations , 1999, IEEE Trans. Knowl. Data Eng..

[20]  Xuemin Lin,et al.  Applying On-line Bitmap Indexing to Reduce Counting Costs in Mining Association Rules , 1999, Inf. Sci..

[21]  Chengqi Zhang,et al.  A fuzzy logic based method to acquire user threshold of minimum-support for mining association rules , 2004, Inf. Sci..

[22]  Ping-Yu Hsu,et al.  Algorithms for mining association rules in bag databases , 2004, Inf. Sci..

[23]  Geoffrey I. Webb,et al.  On detecting differences between groups , 2003, KDD '03.

[24]  Balaji Padmanabhan,et al.  Small is beautiful: discovering the minimal set of unexpected patterns , 2000, KDD '00.

[25]  Masamichi Shimura,et al.  Exceptional Knowledge Discovery in Databases Based on Information Theory , 1996, KDD.