Academia.eduAcademia.edu

Outline

Title
Abstract
Introduction
Experimental Results
Datasets
Experimental Setup
Results
Conclusions
References
All Topics
Computer Science
Machine Learning

Mining Association Rules for Label Ranking

Profile image of CLAUDIO DE SACLAUDIO DE SA

2011, Lecture Notes in Computer Science

Abstract

Recently, a number of learning algorithms have been adapted for label ranking, including instance-based and tree-based methods. In this paper, we continue this line of work by proposing an adaptation of association rules for label ranking based on the APRIORI algorithm. Given that the original APRIORI algorithm does not aim to obtain predictive models, two changes were needed for this achievement. The adaptation essentially consists of using variations of the support and confidence measures based on ranking similarity functions that are suitable for label ranking. Additionally we propose a simple greedy method to select the parameters of the algorithm. We also adapt the method to make a prediction from the possibly conflicting consequents of the rules that apply to an example. Despite having made our adaptation from a very simple variant of association rules for classification, partial results clearly show that the method is making valid predictions. Additionally, they show that it competes well with state-of-the-art label ranking algorithms.

Related papers

Mining Association Rules for Ordinal Data Classification using an Unimodal Model
Alipio Jorge

Some real life problems require the classification of items into naturally ordered classes. Conventional methods, intended for the classification of nominal classes, are traditionally used to deal with these problems where the classes are ordered. This paper proposes an adaptation of association rules for classification intended for multi-class problems where the order relation is not ignored. The theoretical background assumes that the random variable class associated with a given query should follow a unimodal distribution. The adaptation, which uses class association rules (CAR's), is essentially in terms of the output handling, i.e the voting system for the predicted class. The experiments in real datasets are presented. Despite this very simple variant of association rules for classification, the results indicate that the method is making valid predictions and is competitive with state-of-the-art algorithms.

View PDFchevron_right
A novel rule ordering approach in classification association rule mining
Christian Setzkorn

2007

A Classification Association Rule (CAR), a common type of mined knowledge in Data Mining, describes an implicative co-occurring relationship between a set of binary-valued data-attributes (items) and a pre-defined class, expressed in the form of an “antecedent Þ\ Rightarrow (consequent-class” rule. Classification Association Rule Mining (CARM) is a recent Classification Rule Mining (CRM) approach that builds an Association Rule Mining (ARM) based classifier using CARs.

View PDFchevron_right
A new approach to classification based on association rule mining
imam maulana

Classification is one of the key issues in the fields of decision sciences and knowledge discovery. This paper presents a new approach for constructing a classifier, based on an extended association rule mining technique in the context of classification. The characteristic of this approach is threefold: first, applying the information gain measure to the generation of candidate itemsets; second, integrating the process of frequent itemsets generation with the process of rule generation; third, incorporating strategies for avoiding rule redundancy and conflicts into the mining process. The corresponding mining algorithm proposed, namely GARC (Gain based Association Rule Classification), produces a classifier with satisfactory classification accuracy, compared with other classifiers (e.g., C4.5, CBA, SVM, NN). Moreover, in terms of association rule based classification, GARC could filter out many candidate itemsets in the generation process, resulting in a much smaller set of rules than that of CBA.

View PDFchevron_right
Hybrid rule ordering in classification association rule mining
Christian Setzkorn

2008

Abstract Classification Association Rule Mining (CARM) is an approach to classifier generation that builds an Association Rule Mining based classifier using Classification Association Rules (CARs). Regardless of which particular CARM algorithm is used, a similar set of CARs is always generated from data, and a classifier is usually presented as an ordered list of CARs, based on a selected rule ordering strategy. Hence to produce an accurate classifier, it is essential to develop a rational rule ordering mechanism.

View PDFchevron_right
CAR-Miner: An efficient algorithm for mining class-association rules
Loan Nguyen

Expert Systems with Applications, 2013

Classification based on association rules can often achieve higher accuracy than some traditional rule-based methods such as C4.5 and ILA. The right-hand-side part of an association rule is a value of the target (or class) attribute. This study proposes a general algorithm for mining class-association rules based on a variety of interestingness measures. The proposed algorithm uses a tree structure for maintaining the related information of itemsets in the nodes, thus speeding up the process of generation of rules. The proposed algorithm can be easily extended to integrate some measures together for ranking of rules. Experiments are also conducted to show the efficiency of the proposed approach under various settings.

View PDFchevron_right
Association rule evaluation for classification purposes
José María Serrano Chica

Association rules have proved to be useful in building both partial and complete classification models. This paper analyzes alternative measures which could replace confidence in order to evaluate the suitability of a given association rule with respect to the classification problem we try to solve when building a classification model.

View PDFchevron_right
Multi-Label Classification Based on Associations
raed alazaidah

Applied Sciences

Associative classification (AC) has been shown to outperform other methods of single-label classification for over 20 years. In order to create rules that are both more precise and simpler to grasp, AC combines the rules of mining associations with the task of classification. However, the current state of knowledge and the views of various specialists indicate that the issue of multi-label classification (MLC) cannot be solved by any AC method. Since this is the case, adapting or using an AC algorithm to manage multi-label datasets is one of the most pressing issues. To solve the MLC issue, this research proposes modifying the classification based on associations (msCBA) method by extending its capabilities to consider more than one class label in the consequent of its rules and modifying its rules order procedure to fit the nature of the multi-label dataset. The proposed algorithm outperforms several other MLC algorithms from various learning techniques across a variety of performa...

View PDFchevron_right

References (21)

  1. Agrawal, R., Srikant, R.: Fast algorithms for mining association rules. In: Proc. 20th Int. Conf. Very Large Data Bases, VLDB. vol. 1215, p. 487499. Citeseer (1994)
  2. Aiguzhinov, A., Soares, C., Serra, A.P.: A similarity-based adaptation of naive bayes for label ranking. In: Discovery Science (2010)
  3. Azevedo, P.: CAREN-A java based apriori implementation for classification pur- poses (2003)
  4. Brazdil, P., Soares, C., Costa, J.: Ranking Learning Algorithms: Using IBL and Meta-Learning on Accuracy and Time Results. Machine Learning 50(3), 251-277 (2003)
  5. Brin, S., Motwani, R., Ullman, J.D., Tsur, S.: Dynamic itemset counting and im- plication rules for market basket data. Proceedings of the 1997 ACM SIGMOD in- ternational conference on Management of data -SIGMOD '97 pp. 255-264 (1997), http://portal.acm.org/citation.cfm?doid=253260.253325
  6. Cheng, W., Hühn, J., Hüllermeier, E.: Decision tree and instance-based learning for label ranking. In: ICML '09: Proceedings of the 26th Annual International Conference on Machine Learning. pp. 161-168. ACM, New York, NY, USA (2009)
  7. Dekel, O., Manning, C., Singer, Y.: Log-linear models for label ranking. Advances in Neural Information Processing Systems 16 (2003)
  8. Dougherty, J., Kohavi, R., Sahami, M.: Supervised and unsupervised discretization of continuous features. In: MACHINE LEARNING-INTERNATIONAL WORK- SHOP THEN CONFERENCE-. pp. 194-202. MORGAN KAUFMANN PUBLISH- ERS, INC. (1995), http://robotics.stanford.edu/users/sahami/papers-dir/disc.pdf
  9. Fayyad, Irani: Multi-interval discretization of continuous-valued attributes for clas- sification learning. In: International Conference on Machine Learning. pp. 1022- 1027 (1993), http://www.cs.orst.edu/˜bulatov/papers/fayyad-discretization.pdf
  10. Fürnkranz, J., Hüllermeier, E.: Preference learning. KI 19(1), 60-(2005)
  11. Har-Peled, S., Roth, D., Zimak, D.: Constraint classification: a new approach to multiclass classification. In: Proc. of the International Workshop on Algorithmic Learning Theory (ALT). pp. 135-150. Springer-Verlag (2002)
  12. Kemeny, J., Snell, J.: Mathematical Models in the Social Sciences. MIT Press (1972)
  13. Kendall, M., Gibbons, J.: Rank correlation methods (1970)
  14. Lebanon, G., Lafferty, J.D.: Conditional Models on the Ranking Poset. In: NIPS. pp. 415-422 (2002)
  15. Liu, B., Hsu, W., Ma, Y.: Integrating classification and association rule mining. Knowledge Discovery and Data Mining pp. 80-86 (1998), http://www.aaai.org/Library/KDD/1998/kdd98-012.php
  16. Park, J.S., Chen, M.S., Yu, P.S.: An effective hash-based algorithm for min- ing association rules. ACM SIGMOD Record 24(2), 175-186 (May 1995), http://portal.acm.org/citation.cfm?doid=568271.223813
  17. Park, J., Chen, M., Yu, P.: Efficient parallel data mining for as- sociation rules. of the fourth international conference on (1995), http://portal.acm.org/citation.cfm?id=221270.221320
  18. Spearman, C.: The proof and measurement of association between two things. American Journal of Psychology 15, 72-101 (1904)
  19. Thomas, S., Sarawagi, S.: Mining generalized association rules and sequential pat- terns using SQL queries. . Conf. on Knowledge Discovery and Data Mining (1998), http://www.aaai.org/Papers/KDD/1998/KDD98-062.pdf
  20. Todorovski, L., Blockeel, H., Džeroski, S.: Ranking with Predictive Clustering Trees. In: Elomaa, T., Mannila, H., Toivonen, H. (eds.) Proc. of the 13th Euro- pean Conf. on Machine Learning. pp. 444-455. No. 2430 in LNAI, Springer-Verlag (2002) 21. Vembu, S., Gärtner, T.: Label ranking algorithms: A sur- vey. Preference Learning. Springer (2009), http://www-kd.iai.uni- bonn.de/pubattachments/399/VembuG09Book.pdf
  21. Vembu, S., Gärtner, T.: Label Ranking Algorithms: A Survey. In: Johannes Fürnkranz, E.H. (ed.) Preference Learning. Springer-Verlag (2010)

Related papers

Discover Multi-Label Classification using Association Rule Mining
Mehul Parikh

Association rule mining and classification are two major task of data mining. They are attracted wide attention in both research and application area recently. I propose a method for classification rules from multi-label dataset using association rule analysis. Multi label dataset contains multiple class label attribute for predict target variable. We classify that attribute using different approaches like naviye-baies, decision tree, Back propagation, Neural based classification and association rule based classification. Finding association rule from dataset we have to apply various algorithms like Apriori, Fp-Growth, etc. I proposed Fp-Growth algorithm for finding association rule from dataset because of Fp-Growth is an improved algorithm of Apriori and Fp-Growth is more efficient than Apriori. The number of associations present in even moderate sized databases can be, however, very large – usually too large to be applied directly for classification purposes. Therefore, any classi...

View PDFchevron_right
A Novel Rule Weighting Approach in Classification Association Rule Mining
Frans Coenen

Seventh IEEE International Conference on Data Mining Workshops (ICDMW 2007), 2007

. Regardless of which particular CARM algorithm is used, a similar set of CARs is always generated from data, and a classifier is usually presented as an ordered list of CARs, based on a selected rule ordering strategy. Hence to produce an accurate classifier, it is essential to develop a rational rule ordering mechanism. In the past decade, a number of rule ordering strategies have been introduced that can be categorized under three headings: (1) support-confidence, (2) rule weighting, and (3) hybrid. In this paper, we propose an alternative rule weighting scheme, namely CISRW (Class-Item Score based Rule Weighting), and develop a rule weighting based rule ordering mechanism based on CISRW. Subsequently, two hybrid rule ordering strategies are further introduced by combining (1) and CISRW. The experimental results show that the three proposed CISRW based/related rule ordering strategies perform well with respect to the accuracy of classification.

View PDFchevron_right
Learning to rank at query-time using association rules
Pwint Lwin

Proceedings of the 31st …, 2008

Some applications have to present their results in the form of ranked lists. This is the case of many information retrieval applications, in which documents must be sorted according to their relevance to a given query. This has led the interest of the information retrieval community in methods that automatically learn effective ranking functions. In this paper we propose a novel method which uncovers patterns (or rules) in the training data associating features of the document with its relevance to the query, and then uses the discovered rules to rank documents. To address typical problems that are inherent to the utilization of association rules (such as missing rules and rule explosion), the proposed method generates rules on a demand-driven basis, at query-time. The result is an extremely fast and effective ranking method. We conducted a systematic evaluation of the proposed method using the LETOR benchmark collections. We show that generating rules on a demand-driven basis can boost ranking performance, providing gains ranging from 12% to 123%, outperforming the state-of-the-art methods that learn to rank, with no need of time-consuming and laborious pre-processing. As a highlight, we also show that additional information, such as query terms, can make the generated rules more discriminative, further improving ranking performance.

View PDFchevron_right
Generating Classification Association Rules with Modified Apriori Algorithm
Hasan Dag

2006

Association rule mining is a useful and widely used method to extract patterns from large sets of data especially if we deal with basket type data such as customer buying attitudes. It finds all possible relations between data fields , and this property is functional in many research domains; however it happens to be useless in domains like medicine and

View PDFchevron_right
Mining Association Rule in Classified Data for Course Recommender System in E-Learning
Sunita Dol Aher

International Journal of Computer Applications

The ADTree (Alternating Decision Tree) is a supervised classification technique that combines decision trees with the predictive accuracy into a set of classification rules & association rule algorithms are used to show the relationship between data items. Here in this paper we combine these two algorithms & apply it to sample data obtained from Moodle courses of our college for the Course Recommender System which predicts the course selected by the students. First we consider the result using only the association rule then we consider this combined approach. Here we present the advantage of applying the combined approach to Course Recommender System as compare to the result of applying only the association rule algorithm. We found that combined approach works better than only the association rule mining. This combined approach also increases the strength of the rules.

View PDFchevron_right

Related topics

  • Mathematics
  • Computer Science
  • Artificial Intelligence
  • Machine Learning
  • Learning to Rank
  • Association Rule
  • Similarity Function
  • Association Rule Learning
  • Similarity Geometry
    • 580 California St., Suite 400
    San Francisco, CA, 94104
    © 2025 Academia. All rights reserved