R. Agrawal et.al: Fast Discovery of Associattion Rules, Advances in Knowledge Discovery and Data Mining, MIT Press, 1996  Home/Search   Document Not in Database   Summary   Related Articles   Check

....on large web log data. The web log data consists of sequences of URLs requested by different clients bearing different IP addresses. Association rules can be used to decide the next likely web page requests based on significant statistical correlations. In the past, sequential association rules [AS95, AMSTV96] have been used to capture the co occurrence of buying different items in supermarket shopping domains. Episodes were designed to capture significant patterns from sequences of events [MTV95] However, these models were not designed for the prediction task, because they do not specify how to ....

....Association rules [AS94] were proposed to capture the co occurrence of buying different items in a supermarket shopping. It is natural to use association rule generation to relate pages that are most often referenced together in a single server session [SCDT00, CKR98] In the data mining area, [AS95, AMSTV96] proposed sequential association mining algorithms, but these are designed for the discovery of frequent sequential transaction itemsets. They cannot be applied directly for sequential prediction problems because they have to be converted to classifiers first; that 5 is, for any given observation ....

[Article contains additional citation context not shown here]

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, A.I. Verkamo. Fast discovery of association rules. Advances in knowledge discovery and data mining. Pp. 307-328, AAAI/MIT Press, 1996.

....to a real life telecommunications dataset. 1. Introduction Developments in descriptive induction [21, 29] have recently gained much attention of researchers developing rule learning algorithms. These involve mining of association rules (e.g. the APRIORI association rule learning algorithm [1]) clausal discovery (e.g. the CLAUDIEN system [21, 22] subgroup discovery (e.g. the MIDOS [27, 28] and EXPLORA [12] subgroup discovery systems) and other approaches to non classificatory induction aimed at finding interesting patterns in data. In this paper, we consider the task of subgroup ....

....discovery Maximum feature length was set to 8, yielding 276 initial features. An example feature is f99(A) ext number(A,B) prefix(B, 0,4,0,7] meaning that the caller s number starts with 0407. Another feature is f115(A) call date(A,B) call time(A,C) ext number(A,D) prev attempt(B,C,D,[3,1], today,unavailable) meaning that the caller (of the current call) has today tried to reach line 31, which was unavailable. We then set the minimum feature coverage to 20 instances, thus obtaining 138 distinct features. With these features, we use the RSD rule induction algorithm with altered ....

[Article contains additional citation context not shown here]

Rakesh Agrawal, Hiekki Mannila, Ramakrishnan Srikant, Hannu Toivonen, and A. Inkeri Verkamo. Fast discovery of association rules. Advances in knowledge discovery and data mining, pages 307--328, 1996.

....I Introduction Data management today is required of the ability to extract interesting patterns from large and raw data to help decision making, i.e. data mining. Often, patterns are deemed interesting on the basis of passing certain statisti cal tests such as support confidence [AIS93,AMSTV96,AS94]. To an enterprise, however, it remains unclear how such patterns can be used to maximize a busi ness objective. For example, knowing association rules Perfume Lipstick, Perfume Diamond, that are related to Perfume, a store manager wishing to maximize the profit margin still cannot ....

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, A.I. Verkamo. Fast discovery of association rules. Advances in knowledge discovery and data mining, 307-328, AAAI/MIT Press, 1996

....are often used. When data are complex and include categorical attributes, these measures are not easy to de ne [1] Such situations are very usual in medical area. However, association rules method o ers a background to produce clusters without requiring a similarity measure. An association rule [2] is a statement of the form 95 of patients that have gender = male and mediastinum = enlarged also get platelets 2 [100, 600[ 1 , 27 of patients in the database match this rule. This last number is called support of the rule, 95 is the con dence (i.e. the percentage of data that contain the ....

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A. I. Verkamo. Fast discovery of association rules, chapter 12. Advances in Knowledge Discovery and Data Mining. AAAI/MIT Press, 1996.

....leaves, the attribute s values corresponding to the branches leading to a leaf. It permits computation of the global quality of a tree. It corresponds to the difference between the impurity of the root of the tree and the mean impurity of its leaves, this difference being normalized to a value in [0, 1]. The global quality of a tree is equal to 1 if and only if all its leaves are pure and it is equal to 0 if and only if the frequency distributions of D in its root and in all its leaves are identical. The best sub tree for pruning is the one that yields the highest quality pruned tree. C.M. ....

....in a path) Ragel [21] proposes a pre processing method to deal with missing values which temporarily ignored them, without deleting data. This one is based on a search of relevant associations within data to predict missing values. The core of the used algorithm stems from associations rules [1] (we will come back briefly to this method in conclusion) But, with lazy trees, the user has to cope with multi paths and not with a single decision tree. The latter is a graphical structured summary of a data set. Experience in induction tasks shows that this result is a key point of the ....

[Article contains additional citation context not shown here]

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A. I. Verkamo. Fast discovery of association rules, chapter 12. Advances in Knowledge Discovery and Data Mining. AAAI/MIT Press, 1996.

....it to the problem of function prediction of chemical compounds. Matsuda and Motoda et al. 16] presented the algorithm for extracting typical patterns from a directed graph. Their algorithm uses a method called the graph based induction. In the association rule discovery problem, Agrawal et al.[4, 5] developed an algorithm, called Apriori, which is a popular data mining problem. Their algorithm discovers frequent itemsets eciently by using a subset lattice of an itemset. Actually the algorithms [11, 25] described above are based on Apriori. But it is said that the eciency of Apriori slowdowns ....

....all the frequent itemsets without repetition. Sese and Morishita[20] presented the algorithm discovering optimal itemsets, based on the set enumeration tree and a method of merging occurrence lists of each itemset. In addition to these works, there are many works that complement Apriori algorithm [3, 4, 5, 9, 14, 20] Our results generalizes the techniques of [9] and [20] above, and thus can be regarded as a tree counter part of the second generation of the association mining techniques. 1.2 Organization The rest of this paper is organized as follows. In Section 2, we prepare basic notions and de nitions. In ....

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, A. I. Verkamo, Fast Discovery of Association Rules, Advances in Knowledge Discovery and Data Mining, Chapter 12, AAAI Press / The MIT Press, 1996.

....topics we added on the fly as needed, without re classifying documents inserted in the past. This fact has a negative impact on classifier training. 2 Itemsets and Apriori Algorithm The apriori algorithm (Agrawal et al. is an efficient algorithm for knowledge mining in form of association rules [2]. We have recognized its convenience for document categorization. The original apriori algorithm is applied to a transactional database of market baskets. In our case, instead of a market basket, we work with the basket of significant terms occurring in a text document and the transactional ....

Agrawal et al.: Advances in Knowledge Discovery and Data Mining, MIT Press 1996, pp. 307-328

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A.I. Verkamo. Fast discovery of association rules. Advances in Knowledge Discovery and Data Mining, pp 307--328. AAAI/MIT Press, 1996.

....the input S is seen asan unordered collection ofalarm s.In Step 4, recognition ,for each candidate set the number of satisfying alarms is computed and compared to the frequency threshold. In Step 5, building , new candidate sets are constructed such that every subset of a candidate set is frequent [31]. This basic algorithm can be modi ed to take into account, e.g. the net work topology, the types of network elements, or an alarm type hierarchy. For instance, episodes can be required to consist of alarms from network elements whose distance is small in the network topology. A simple way to ....

R. Agrawal, H. Mannila, R.Srikant, H. Toivonen, and A. I. Verkamo, Fast discovery of associa - tionrules. InU. M.Fayyad, G.Piatetsky - S hapiro, P.Smyth, andR. Uthurusamy (eds.), Advances in Knowledge Discovery and Data Mining , AAAI Press, Menlo Park, California, pp. 307328, 1996.

No context found.

Agrawal, 1L, Mannila, H., Srikant, R. et al. Fast discovery of association rules. Advances in Knowledge Discovery and Data Mining. AAAI Press, Menlo Park, California, USA, 1996, pp. 307-328.

....patterns, they have to investigate specific patterns to find cardinalities of subgroups or significance of deviations, etc. Typically, the patterns, whose frequencies are needed, are conjunctions of atomic patterns. A prime example is given by the frequent set concept underlying association rules [2, 3]. Moreover, the patterns defined for correlation [6, 7] causality [18] sequential patterns [4] episodes [13] constrained frequent sets [11, 14, 19] long patterns [1, 5] closed sets [16] and many other important data mining tasks have the same basic form. In all these cases, we have ....

....containing about 5000 transactions of about 200 distinct types of telecommunications network alarms. For proprietary reasons, we cannot describe this data set further. 2. regular synthetic data set, which is a synthetic data set generated using the program developed at IBM Almaden Research Center [3]. The exact number of transactions is not important, because the key parameter is the number of pages p. In our experimentation, p varies from 200 to 50 000, and the number of items is k = 1000. 3. skewed synthetic data set, which is a synthetic data set that has skewed seasonal behavior. ....

R. Agrawal, H. Mannila, et al. Fast discovery of association rules. Advances in Knowledge Discovery and Data Mining, pp 307--328. AAAI/MIT Press, 1996.

No context found.

R. Agrawal et.al: Fast Discovery of Associattion Rules, Advances in Knowledge Discovery and Data Mining, MIT Press, 1996

No context found.

Agrawal, R., Mannila, H., Srikant, R., Toivonen, H., Verkamo, A. I.: Fast Discovery of Association Rules, Advances in Knowledge Discovery and Data Mining (U. Fayyad, et al, Eds.), AAAI Press, Menlo Park, CA, 1996.

No context found.

Agrawal R., Mannila H., Srikant R., Toivonen H., and Verkamo A. Fast Discovery of Association Rules. Advances in Knowledge Discovery and Data Mining, pp.307-- 328, 1996.

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A.I. Verkamo, Fast discovery of association rules, Advances in Knowledge Discovery and Data Mining (1996), 307--328.

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A. Verkamo. Fast discovery of association rules. Advances in Knowledge Discovery and Data Mining, 1996.

No context found.

Agrawal, R., Mannila, H., Srikant, R., Toivonen, H., and Verkamo, A.I., Fast Discovery of Association Rules, Advances in Knowledge Discovery and Data Mining, Chapter 12, AAAI/MIT Press, 1995

No context found.

R. Agrawal et.al: Fast Discovery of Associattion Rules, Advances in Knowledge Discovery and Data Mining, MIT Press, 1996

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen and A. I. Verkamo, Fast Discovery of Association Rules, Advances in Knowledge Discovery and Data Mining, edited by U. M. Fayyad et al, AAAI/MIT Press, pp. 307-328, 1996.

No context found.

Rakesh Agrawal, Hiekki Mannila, Ramakrishnan Srikant, Hannu Toivonen, and A. Inkeri Verkamo. Fast discovery of association rules. Advances in knowledge discovery and data mining, pages 307--328, 1996.

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A. I. Verkamo. Advances in knowledge discovery and data mining, chapter 12,. In Fast Discovery of Association Rules. AAAI/MIT Press, MA, 1995.

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A. I. Verkamo, Fast discovery of association rules, Advances in Knowledge Discovery and Data Mining, Chap. 12, MIT Press, 307-328, 1996.

No context found.

Rakesh Agrawal, Heikki Mannila, Ramakrishnan Srikant, Hannu Toivonen and A. Inkery Verkamo. Advances in Knowledge Discovery and Data Mining, chapter Fast Discovery of Association Rules, pages 307--328. MIT Press, 1996.

No context found.

Agrawal R., Manilia H., Srikant R., Toivonen H., Verkamo A.: Fast Discovery of Association Rules, Advances in Knowledge Discovery and Data Mining, 307-328, AAAI press, Menlo Park, CA, 1996. Abbreviated presentation of Apriori algorithm and its derivatives.

No context found.

R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, and A.I. Verkamo. Fast Discovery of Association Rules. Advances in Knowledge Discovery and Data Mining. AAAI/MIT Press, 1996, pp. 307-328.

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