G. Webb, Efficient search for association rules. In: Proceedings of ACM SIGKDD, 2000: 99-107.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the basic association paradigm in two ways: they provide performance improvements based on a new method of enumerating large itemsets and additionally propose the notion of implication rules as an alternative to association rules, introducing the notion of conviction. Bayardo et al. 4] and Webb [20] propose branch and bound algorithms for searching the space of possible associations. Their algorithms apply pruning rules that do not rely solely on support (as in the case of a priori algorithms) Cohen et al. 7] propose an algorithm for fast mining of associations with high confidence without ....

G. Webb, Efficient Search for association rules, KDD, 2000.

....as a request for the additional items that have sold together with sausage in order to make it highly likely that mustard will also be sold. 2.4. 1 Representation A training set is a finite set of records where each record is an element to which Boolean predicates called conditions are applied [27]. A rule can be expressed C A , where antecedent A implies consequent C. The antecedent is often called the left handside or LHS and the consequent the right hand side or RHS [28] Initial association rule definitions [12] only allowed one variable in the consequent, however this variable might ....

....A implies consequent C. The antecedent is often called the left handside or LHS and the consequent the right hand side or RHS [28] Initial association rule definitions [12] only allowed one variable in the consequent, however this variable might differ form rule to rule. More recent publications [27, 29] allow many target attributes to be discovered as the consequent in rules. The association rule problem is interested in discovery of rules that meet the following constraints [12] 13 . Syntactic constraints: Constraints that involve restrictions on the items that can appear in a rule. ....

[Article contains additional citation context not shown here]

Webb, G.I. Efficient Search for Association Rules. in The Sixth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2000. 58

....Furthermore, the original CN2 contains a post processing step for pruning the resulting rules. 10 Association rule mining is actually not a typical candidate for the Cheese framework, because the most widely used algorithm Apriori is not a search in our sense. However, we adopted an idea from [Web00], which introduces a direct search method for association rules: ffl Graph: antecedent consequent top down ffl Evaluator: confidence support threshold evaluator ffl Subspace Evaluator: optimistic confidence support pruning ffl Local Reductor: none ffl Global Reductor: none ffl Search Mode: ....

....uses this information to store only those rules with an evaluation of 1.0. Thus, a model of two evaluation values (confidence support) is implemented as is common in the association rule framework. Within the subspace evaluator we implemented some (yet not all) of the pruning rules described in [Web00]. This yields an optimistic pruning strategy by not expanding certain regions of the subspace any further, unless certain criteria are met. Further implementations: The two examples listed above should suffice to illustrate the basic principle. Besides that we also implemented a stochastic ....

G.I. Webb. Efficient search for association rules. In Knowledge Discovery and Data Mining, pages 99--107, 2000. 13

....be applied for effectively reducing the number of candidates (or can only reduce them at the expense of loosing interesting patterns) the level wise search pioneered by Apriori might in fact not be a suitable choice. In particular if the search is performed in memory, alternative search strategies [5, 16] might be considered because of their more flexible pruning strategies. It is an open research problem which strategy is more appropriate for multi relational data mining systems like Warmr. 6 Related Work The problem of detecting temporal change in frequent patterns by grouping objects over ....

G. I. Webb. Efficient search for association rules. In Proceedings of the 6th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD-

....of records defies any form of manual analysis, rendering the use of algorithms for pattern analysis indispensable. Our approach to inducing rules combines the strength of association rule mining [AMS 96] with a systematic search for rules under strong pruning techniques [Rym93, Web93, Web95, Web00] Our analysis is similar to the general framework for rule induction proposed by [Web00] but with important modifications (Section 4) We show how specific settings for the modified framework provide a principled approach to the construction of accurate rules correlating computer events. Our ....

....analysis indispensable. Our approach to inducing rules combines the strength of association rule mining [AMS 96] with a systematic search for rules under strong pruning techniques [Rym93, Web93, Web95, Web00] Our analysis is similar to the general framework for rule induction proposed by [Web00] but with important modifications (Section 4) We show how specific settings for the modified framework provide a principled approach to the construction of accurate rules correlating computer events. Our experiments show the effects of varying several parameters on the quality of the induced ....

G. I. Webb. Efficient search for association rules. In In Proceedings of the Sixth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 99--107, 2000.

....from all of the frequent itemsets, is relatively straightforward, but it can still be very expensive when solving real world problems. We have only briefly described the most basic concepts of association rule discovery. For more detailed information, see related technical publications [1] 2] 4][12][13] 3. ALGORITHMS In this section we describe the software implementations of the association rule algorithms used in our experiments. The five algorithms evaluated were Apriori, Charm, FP growth, Closet and MagnumOpus. We provide references to articles describing the details of the algorithm ....

....and Closet take transactional data in the form of one row for each single complete transaction, with the number of items in the transaction followed by a list of items. These implementations of FP growth and Closet only generate the frequent itemsets, and not the association rules. MagnumOpus: MO [12] is the command line application shipped with the beta release of MagnumOpus1.2, a commercial system for association rule discovery. The main unique technique used in MagnumOpus is the search algorithm based on OPUS [11] a systematic search method with pruning. It considers the whole search ....

[Article contains additional citation context not shown here]

Webb, G.I. Efficient search for association rules. In Proceedings of the Sixth ACM-SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY: ACM, 99-107.

....of this first step. The second step is relatively straightforward, but it can still be very expensive when solving real world problems. We have only briefly described the most basic concepts of association rule discovery. For more detailed information, see related technical publications [1] 2][9][10] 3. ALGORITHMS In this section we describe the software implementations of the association rule algorithms used in our experiments. The five algorithms evaluated were Apriori, Charm, FP growth, Closet and MagnumOpus. We started the experiments several months ago and published preliminary ....

....We received this implementation on September 21, 2000. Both FP growth and Closet take transactional data in the form of one row for each single complete transaction. These implementations of FP growth and Closet only generate the frequent itemsets, and not the association rules. MagnumOpus: MO [9] is the command line application shipped with the beta release of MagnumOpus1.2, a commercial system for association rule discovery. The main unique technique used in MagnumOpus is the search algorithm based on OPUS [8] a systematic search method with pruning. It considers the whole search space, ....

[Article contains additional citation context not shown here]

Webb, G.I. Efficient search for association rules. In Proceedings of the Sixth ACM-SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY: ACM, 99-107.

....of mining rules for all target classes, it only mines rules for one target item. It differs from CBA RG in that it will only mine a number of rules within a certain range. For further details, see [12] A different rule mining algorithm that produces rulesets of constrained size is presented in [18]. 3 Recommendation using our Mining Algorithm We now describe how our algorithm to mine association rules may be used for collaborative recommendation. The same mining process may be used to mine a certain number of rules for each user article for both user associations and article associations. ....

G. I. Webb. Efficient search for association rules. In Proc. KDD-2000, Boston, August 2000, to appear.

No context found.

G. Webb, Efficient search for association rules. In: Proceedings of ACM SIGKDD, 2000: 99-107.

No context found.

G. Webb, Efficient search for association rules. In: Proceedings of ACM International Conference on Knowledge Discovery and Data Mining, 2000: 99-107.

No context found.

G. Webb, Efficient search for association rules. In: Proceedings of ACM International Conference on Knowledge Discovery and Data Mining, 2000: 99-107.

No context found.

G. Webb, Efficient search for association rules. In: Proceedings of ACM SIGKDD, 2000: 99-107.

No context found.

G. I. Webb. Efficient search for association rules. In R. Ramakrishnan, S. Stolfo, R. Bayardo, and I. Parsa, editors, Proceedinmgs of the 6th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD-00), pages 99--107, N. Y., Aug. 20--23 2000. ACM Press.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC