Klementinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A. I. - Finding interesting rules from large sets of discovered association rules. In Third International Conference on Information and Knowledge Management -- (1994) 401 -- 407  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the information, the more abundant the data. The basic objective of data mining techniques is to extract useful patterns (i.e. actionable knowledge) from these large databases. State of the art systems, however, still discover too many redundant patterns. In the past few years many attempts [1] [2], 3] 4] have been made to identify the best criteria for evaluating the interestingness of patterns, as only few among them are really relevant to the decision makers. So far, two main classes of approaches to the evaluation of interestingness have been identified: the techniques belonging to ....

....the techniques belonging to the first group are based on objective measures of interestingness, the others on subjective ones. Objective measures, such as confidence, support, strength, simplicity [5] 6] 7] 8] are focused on the statistical strength of a pattern, while subjective measures [2], 9] are based on the assumption that the interestingness of a pattern strongly depends on the personal expectations of the analyst. In this paper our main concern is to define an effective interestingness measure apt to be visualized, so that final users of the system can easily understand it. ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H. & Verkamo, A.I., Finding interesting rules from large sets of discovered association rules. CIKM-94, pp. 401-407, 1994.

....the number of patterns rules discovered under the support model can be very large. Many post processing techniques have been developed to reduce the number of discovered patterns into a manageable size while preserving the discovered knowledge as much as possible. Human interaction is involved in [12, 22, 23] to specify the interestingness or beliefs to guide the process while others [13, 14] focused on reducing redundant information possessed by the discovered rules. It is clear that these post processing techniques are typically used as an additional pruning step after the normal mining procedure ....

M. Klemetinen, H. Mannila, P. Ronkainen, H. Toivonen, and A. Verkamo. Finding interesting rules from large sets of discovered association rules. Proc. CIKM, 1994.

....(i.e. 1) towards its significance, regardless of its likelihood of occurrence. Intuitively, the assessment of significance of a pattern in a sequence should take into account the expectation of pattern occurrence (according to some prior knowledge) Recently, many research has been proposed [1, 3, 5, 6, 8, 9, 10, 11, 12, 15] towards this objective. We will furnish an overview in the next section. In this paper, a new model is proposed to characterize the class of so called surprising patterns (instead of frequent patterns) We will see that our model not only has a solid theoretical foundation but also allows an ....

M. Klemetinen, H. Mannila, P. Ronkainen, H. Toivonen, and A. Verkamo. Finding interesting rules from large sets of discovered association rules. Proc. CIKM, 1994.

....powerful regularities in binary data. An association rule is an expression of the form X Y, where X and Y are sets of items. The intuitive meaning of such a rule is that in the rows of the database where the attributes in X have value true, also the attribute Y has value true with high probability [12]. The problem is to mine for rules that satisfy user specified minimum support and minimum confidence. There are hundreds of association rules in a given data set depending on its size and complexity. The process of mining for such rules is called Association Rule Mining. Much of the genomic ....

Klemttinen, M., Mannila, H., Ronkainen, P., Toivonen, H., Verkamo, A.I. 1994 Finding Interesting Rules from Large Sets of Discovered Association Rules, Third Int'l Conf. on Information and Knowledge Management (CIKM'94): 401 407

....and electronic commerce. The association rule model was introduced by Agrawal, Imielinski, and Swami [5] Starting with the pioneering work in [5] the association rule problem and its variations have been studied extensively by researchers. Several variations of the association rule problem [4, 8, 10, 19] have been proposed which can provide more interesting rules than the support confidence framework. In addition, a number of methods have been discussed in the literature which extend the binary association rule problem to related scenarios such as quantitative association rules, generalized ....

M. Klementtinen, H. Mannila, P. Ronkainen, H. Toivonen, A. I. Verkamo. Finding Interesting Rules from Large Sets of discovered association rules. CIKM Conference Proceedings, pages 401--407, 1994.

....also contain a large amount of redundancy [8] Past research in dealing with this problem can be described with the following approaches: a) Discover all rules first and then allow the user to query and retrieve those he she is interested in. The representative approach is that of templates [3]. This approach lets the user to specify what rules he she is interested as templates. The system then uses the templates to retrieve the rules that match the templates from the set of discovered rules. b) Use constraints to constrain the mining process to generate only relevant rules. 12] ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A. I. "Finding interesting rules from large sets of discovered association rules," CIKM-1994.

....performed to remove those redundant and or insignificant rules. Most existing approaches to finding subjectively interesting association rules ask the user to explicitly specify what types of rules are interesting and uninteresting. The system then generates or retrieves those matching rules. [10] proposes a template based approach. In this approach, the user specifies interesting and uninteresting association rules using templates. A template describes a set of rules in terms of items occurred in the conditional and the consequent parts. The system then retrieves the matching rules from ....

....rule mining algorithm in [28] Those redundant and or insignificant rules are removed using the pruning technique in [14] objective interestingness analysis) Since there is no existing technique that is able to perform our task, we could not carry out a comparison. Most existing methods [10, 7, 8, 18, 20, 29] only produce conforming rules but not unexpected rules. Although the system described in [23, 24] produces unexpected association rules, it is not an interactive post analysis system, and it does not handle RPC and GI specifications. As the proposed technique deals with subjective ....

. Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. 1994. "Finding interesting rules from large sets of discovered association rules." Proceedings of the Third International Conference on Information and Knowledge Management, 1994, pp. 401-407.

....rules to gain a good understanding of the domain is one of the important phases of the KDD process. It usually requires the user to browse a large set of discovered rules. Typical techniques (commonly called post processing techniques) that assist the user in the process include templates [9], expectations [11, 21] summarization [13] and visualization [20] In this paper, we focus on using the web to help the user to interpret a set of association rules. Finding interesting useful knowledge from a set of association rules is a particularly hard problem as the number of rules is ....

. Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. "Finding interesting rules from large sets of discovered association rules." Proceeding of the International Conference on Information and Knowledge Management (CIKM-1994), 1994.

....technique is applied differently depending on the context. Experiment results and practical applications show that the proposed technique is very effective and efficient. 2. RELATED WORK In the past few years, a number of techniques were proposed to deal with the problem of too many rules [21, 23, 12, 14, 9, 19, 1]. The main idea of these techniques is to use the user s knowledge or statistical measures to remove those uninteresting rules. Our work is different. We aim to organize and summarize the rules so that the user can browse through them easily and effectively. 9] proposes a template based approach ....

....many rules [21, 23, 12, 14, 9, 19, 1] The main idea of these techniques is to use the user s knowledge or statistical measures to remove those uninteresting rules. Our work is different. We aim to organize and summarize the rules so that the user can browse through them easily and effectively. [9] proposes a template based approach for finding interesting rules. This approach first asks the user to specify what rules he she wants. The system then finds those matching rules. The technique assumes that the user knows exactly what he she is looking for. However, in many situations, the user ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. "Finding interesting rules from large sets of discovered association rules." CIKM-1994.

....they also contain a large amount of redundancy [8] Past research in dealing with this problem can be described with the following approaches: a) Discover all rules first and then allow the user to query and retrieve those he she is interested in. The representative approach is that of templates [3]. This approach lets the user to specify what rules he she is interested as templates. The system then uses the templates to retrieve the rules that match the templates from the set of discovered rules. b) Use constraints to constrain the mining process to generate only relevant rules. 12] ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A. I. "Finding interesting rules from large sets of discovered association rules," CIKM, 1994.

....been studied by many researchers in data mining. However, to the best of our knowledge, there is no existing work that tackles the problem from a knowledge representation point of view. In the interestingness research of data mining, a number of techniques (e.g. Piatesky Shapiro Matheus 1994; Klemetinen et al. 1994; Silberschatz Tuzhilin 1996; Liu Hsu 1996; Padmanabhan Tuzhilin 1998) have been proposed to help the user find interesting rules from a large number of discovered rules. The main approaches are: 1) using some interestingness measures to filter out those uninteresting rules; and (2) using ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H. and Verkamo, A. 1994. Finding interesting rules from large sets of discovered association rules. CIKM-94.

....That is why, subjective interestingness may be biased and may vary with different users. Different methods have been proposed to capture interesting rules subjectively. The main idea is to impose user s own belief about the domain. Users usually apply their knowledge in terms of rule templates [6] and then try to match the template by scanning the data satisfying some threshold parameters. This approach may be suitable to justify a particular user s own belief system but may fail to discover some surprising rules that they even don t know. One potential problem is that user s subjective ....

H. Mannila M. Klemettinen. Finding interesting rules from large sets of discovered association rules. In Third Intl. Conference on Information and Knowledge Management (CIKM), 1994.

....makes association rule mining effective and practical for data sets whose items are highly correlated. The user can now obtain a complete picture of the domain without being overwhelmed by a huge number of rules. 2. Related Work The problem of too many rules has been studied by many researchers. [8] proposed an approach to allow the user to specify what he she wants to see using templates. The system then retrieves those match rules from the set of discovered rules. This method, however, does not prune those insignificant rules and does not provide a summary of the discovered rules. In ....

....does not prune those insignificant rules and does not provide a summary of the discovered rules. In subjective interestingness research in data mining, 22, 11, 12, 19] proposed a number of methods for finding unexpected rules. Instead of asking the user to specify what he she wants to see as in [8], these approaches ask the user Pruning Significant rules Summarization DS rules Discovered large rules Non DS rules to specify his her existing knowledge about the domain. The system then finds those unexpected rules by comparing the user s knowledge with the discovered rules. Again, these ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. "Finding interesting rules from large sets of discovered association rules." CIKM-1994.

....rule set has 770 rules) Without the proposed system, it would be very hard for us to analyze these large numbers of rules. 5. Related Work Traditionally, a query based approach is used to help the user identify or generate interesting rules. The approach takes many forms, e.g. templates [6], M SQL [5] DMQL [4] and action hierarchy [1] Although query languages can be quite different, a query basically defines a set of rules of a certain type (or constraints on the rules to be found) To execute a query means to find all rules that satisfy the query. We believe that the ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. "Finding interesting rules from large sets of discovered association rules." CIKM-94, 1994, pp. 401-407.

....Some automated assistance is needed. Identifying interesting rules from a set of discovered rules is not a simple task because a rule could be interesting to one user but not interesting to another. The interestingness of a rule is essentially subjective (e.g. PiateskyShapiro et al. 1994b; Klemetinen et al. 1994; Silberschatz and Tuzhilin 1996; Liu and Hsu 1996) because it depends on the user s existing concepts about the domain, and his her interests. There is also an objective aspect of interestingness, which is not studied here. Interested readers, please refer to (Major and Mangano 1993; Silberschatz ....

....Matwin 1993) Clearly, they are different from our work, which aims to help the user analyze the discovered rules in order to identify those interesting ones. In data mining, subjective interestingness (e.g. PiateskyShapiro and Matheus 1994a; Piatesky Shapiro et al. 1994b; Major and Mangano 1993; Klemetinen et al. 1994) has long been identified as an important problem. PiateskyShapiro and Matheus 1994a) studied the issue in a health care application. The system (called KEFIR) analyzes the health care information to uncover interesting deviations. However, KEFIR does not perform rule comparison. Its approach is ....

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. 1994. Finding interesting rules from large sets of discovered association rules. Proceedings of the Third International Conference on Information and Knowledge Management, 401-407.

.... (KDD) has become an important field in recent years to address the need for analyzing data contained in very large database [13] Among discovering many kinds of knowledge in large databases, association rule mining has attracted great attention in database research communities in recent years [2, 4, 25, 36, 33, 20, 39, 40, 42, 30, 41, 31, 8]. Association rule mining is a form of data mining to discover interesting relationships among attributes in those data. The discovered rules may help marketing, decision making, and business management. An example of such a rule might be that 98 of customers that purchase tires and auto ....

....was first introduced in [2] Since then, efficient association mining mechanism in large databases and its extensions to different domains have been the subject of many studies. These studies cover a broad spectrum of topics including: 1) fast algorithms based on the level wise Apriori framework [4, 25] and its variations, including partitioning [36, 33] and sampling [42] 2) incremental updating and parallel algorithms [11, 34, 16] 3) mining of generalized and multi level rules [39, 20] 4) mining of quantitative and multi dimensional rules [40, 14, 30, 27, 23] 5) mining long patterns ....

[Article contains additional citation context not shown here]

M. Klemettinen, H. Mannila, P. Ronkainen, H. Toivonen, and A.I. Verkamo. Finding interesting rules from large sets of discovered association rules. In Proc. 3rd Int. Conf. Information and Knowledge Management, pages 401--408, Gaithersburg, Maryland, Nov. 1994.

....pruning and sorting criteria for association and episode rules are rule confidence, frequency, and statistical significance. Rules can be pruned by setting thresholds on these properties. Examples of the effect of threshold value based pruning with alarm data are presented in Figure 3. Templates [10], pattern expressions that describe the form of rules that are to be selected or rejected, are one way of focusing the view to a large space of rules. A template is an expression A1 ; Ak ) Ak 1 ; A l , where each A i is either an attribute name, a class name, or an expression C ....

M. Klemettinen, H. Mannila, P. Ronkainen, H. Toivonen, and I. Verkamo. Finding interesting rules from large sets of discovered association rules. In Proceedings of the Third Int'l Conference on Information and Knowledge Management (CIKM'94), pp. 401 -- 407, Gaithersburg, MD, November 1994. ACM.

No context found.

Klementinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A. I. - Finding interesting rules from large sets of discovered association rules. In Third International Conference on Information and Knowledge Management -- (1994) 401 -- 407

No context found.

Klementinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A. I. - Finding interesting rules from large sets of discovered association rules. In 3 Int. Conf. on Information and Knowledge Management -- (1994)

No context found.

Klementtinen M., Mannila H., Ronkainen P., Toivonen H., and Verkamo A. I. Finding interesting rules from large sets of discovered association rules. Proceedings of the CIKM 1994.

No context found.

Klementtinen M., Mannila H., Ronkainen P., Toivonen H., and Verkamo A. I. Finding interesting rules from large sets of discovered association rules. Proceedings of the Conference on Information and Knowledge Managements. Gaithersburg, MD, USA 28 Nov. 2 Dec. 1994.

No context found.

Klemetinen, M., Mannila, H., Ronkainen, P., Toivonen, H., and Verkamo, A.I. "Finding interesting rules from large sets of discovered association rules." CIKM-94, 1994.

No context found.

Klementtinen M., Mannila H., Ronkainen P., Toivonen H., and Verkamo A. I. Finding interesting rules from large sets of discovered association rules. Proceedings of the Conference on Information and Knowledge Managements. Gaithersburg, MD, USA 28 Nov. 2 Dec. 1994.

No context found.

Klementtinen M., Mannila H., Ronkainen P., Toivonen H., and Verkamo A. I. "Finding interesting rules from large sets of discovered association rules." Proceedings of the CIKM 1994.

No context found.

M. Flemettinen, H. Mannila, P. Ronkainen, H. Toivonen, A. I. Verkamo, Finding Interesting Rules from Large Sets of Discovered Association Rules, the 3rd International Conference on Information and Knowledge Management, pp.401407, 1994.

First 50 documents

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