TKQ: Top-K Quantitative High Utility Itemset Mining

Mourad Nouioua; Philippe Fournier-Viger; Wensheng Gan; Youxi Wu; Jerry Chun Wei Lin; Farid Nouioua

doi:10.1007/978-3-030-95408-6_2

TKQ: Top-K Quantitative High Utility Itemset Mining

Mourad Nouioua, Philippe Fournier-Viger^*, Wensheng Gan, Youxi Wu, Jerry Chun Wei Lin, Farid Nouioua

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

6 Scopus citations

Abstract

High utility itemset mining is a well-studied data mining task for analyzing customer transactions. It consists of finding the sets of items purchased together that yield a profit that is greater than a minutil threshold, set by the user. To find more precise patterns with purchase quantities, that task was recently generalized as high utility quantitative itemset mining. But an important drawback of current algorithms is that finding an appropriate minutil value is not intuitive and can greatly influence the output. A too small minutil value may lead to very long runtimes and finding millions of patterns, while a too high value, may result in missing many important patterns. To address this issue, this paper redefines the task as top-k quantitative high utility itemset mining and proposes a novel algorithm named TKQ (Top K Quantitative itemset miner), which let the user directly specify the number k of patterns to be found. The algorithm includes three strategies to improve its performance. Experiments on benchmark datasets show that TKQ has excellent performance.

Original language	English
Title of host publication	Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings
Editors	Bohan Li, Lin Yue, Jing Jiang, Weitong Chen, Xue Li, Guodong Long, Fei Fang, Han Yu
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	16-28
Number of pages	13
ISBN (Print)	9783030954079
DOIs	https://doi.org/10.1007/978-3-030-95408-6_2
State	Published - 2022
Externally published	Yes
Event	17th International Conference on Advanced Data Mining and Applications, ADMA 2021 - Sydney, Australia Duration: 2 Feb 2022 → 4 Feb 2022

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13088 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	17th International Conference on Advanced Data Mining and Applications, ADMA 2021
Country/Territory	Australia
City	Sydney
Period	2/02/22 → 4/02/22

Bibliographical note

Publisher Copyright:
© 2022, Springer Nature Switzerland AG.

Keywords

High utility itemset mining
Quantitative itemsets
Top-k Pattern mining

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-95408-6_2

Cite this

Nouioua, M., Fournier-Viger, P., Gan, W., Wu, Y., Lin, J. C. W., & Nouioua, F. (2022). TKQ: Top-K Quantitative High Utility Itemset Mining. In B. Li, L. Yue, J. Jiang, W. Chen, X. Li, G. Long, F. Fang, & H. Yu (Eds.), Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings (pp. 16-28). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13088 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-95408-6_2

Nouioua, Mourad ; Fournier-Viger, Philippe ; Gan, Wensheng et al. / TKQ : Top-K Quantitative High Utility Itemset Mining. Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings. editor / Bohan Li ; Lin Yue ; Jing Jiang ; Weitong Chen ; Xue Li ; Guodong Long ; Fei Fang ; Han Yu. Springer Science and Business Media Deutschland GmbH, 2022. pp. 16-28 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{edd55f8d9636490b907ad3d318ddeb2d,

title = "TKQ: Top-K Quantitative High Utility Itemset Mining",

abstract = "High utility itemset mining is a well-studied data mining task for analyzing customer transactions. It consists of finding the sets of items purchased together that yield a profit that is greater than a minutil threshold, set by the user. To find more precise patterns with purchase quantities, that task was recently generalized as high utility quantitative itemset mining. But an important drawback of current algorithms is that finding an appropriate minutil value is not intuitive and can greatly influence the output. A too small minutil value may lead to very long runtimes and finding millions of patterns, while a too high value, may result in missing many important patterns. To address this issue, this paper redefines the task as top-k quantitative high utility itemset mining and proposes a novel algorithm named TKQ (Top K Quantitative itemset miner), which let the user directly specify the number k of patterns to be found. The algorithm includes three strategies to improve its performance. Experiments on benchmark datasets show that TKQ has excellent performance.",

keywords = "High utility itemset mining, Quantitative itemsets, Top-k Pattern mining",

author = "Mourad Nouioua and Philippe Fournier-Viger and Wensheng Gan and Youxi Wu and Lin, \{Jerry Chun Wei\} and Farid Nouioua",

note = "Publisher Copyright: {\textcopyright} 2022, Springer Nature Switzerland AG.; 17th International Conference on Advanced Data Mining and Applications, ADMA 2021 ; Conference date: 02-02-2022 Through 04-02-2022",

year = "2022",

doi = "10.1007/978-3-030-95408-6\_2",

language = "English",

isbn = "9783030954079",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "16--28",

editor = "Bohan Li and Lin Yue and Jing Jiang and Weitong Chen and Xue Li and Guodong Long and Fei Fang and Han Yu",

booktitle = "Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings",

address = "Germany",

}

Nouioua, M, Fournier-Viger, P, Gan, W, Wu, Y, Lin, JCW & Nouioua, F 2022, TKQ: Top-K Quantitative High Utility Itemset Mining. in B Li, L Yue, J Jiang, W Chen, X Li, G Long, F Fang & H Yu (eds), Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13088 LNAI, Springer Science and Business Media Deutschland GmbH, pp. 16-28, 17th International Conference on Advanced Data Mining and Applications, ADMA 2021, Sydney, Australia, 2/02/22. https://doi.org/10.1007/978-3-030-95408-6_2

TKQ: Top-K Quantitative High Utility Itemset Mining. / Nouioua, Mourad; Fournier-Viger, Philippe; Gan, Wensheng et al.
Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings. ed. / Bohan Li; Lin Yue; Jing Jiang; Weitong Chen; Xue Li; Guodong Long; Fei Fang; Han Yu. Springer Science and Business Media Deutschland GmbH, 2022. p. 16-28 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13088 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - TKQ

T2 - 17th International Conference on Advanced Data Mining and Applications, ADMA 2021

AU - Nouioua, Mourad

AU - Fournier-Viger, Philippe

AU - Gan, Wensheng

AU - Wu, Youxi

AU - Lin, Jerry Chun Wei

AU - Nouioua, Farid

PY - 2022

Y1 - 2022

N2 - High utility itemset mining is a well-studied data mining task for analyzing customer transactions. It consists of finding the sets of items purchased together that yield a profit that is greater than a minutil threshold, set by the user. To find more precise patterns with purchase quantities, that task was recently generalized as high utility quantitative itemset mining. But an important drawback of current algorithms is that finding an appropriate minutil value is not intuitive and can greatly influence the output. A too small minutil value may lead to very long runtimes and finding millions of patterns, while a too high value, may result in missing many important patterns. To address this issue, this paper redefines the task as top-k quantitative high utility itemset mining and proposes a novel algorithm named TKQ (Top K Quantitative itemset miner), which let the user directly specify the number k of patterns to be found. The algorithm includes three strategies to improve its performance. Experiments on benchmark datasets show that TKQ has excellent performance.

AB - High utility itemset mining is a well-studied data mining task for analyzing customer transactions. It consists of finding the sets of items purchased together that yield a profit that is greater than a minutil threshold, set by the user. To find more precise patterns with purchase quantities, that task was recently generalized as high utility quantitative itemset mining. But an important drawback of current algorithms is that finding an appropriate minutil value is not intuitive and can greatly influence the output. A too small minutil value may lead to very long runtimes and finding millions of patterns, while a too high value, may result in missing many important patterns. To address this issue, this paper redefines the task as top-k quantitative high utility itemset mining and proposes a novel algorithm named TKQ (Top K Quantitative itemset miner), which let the user directly specify the number k of patterns to be found. The algorithm includes three strategies to improve its performance. Experiments on benchmark datasets show that TKQ has excellent performance.

KW - High utility itemset mining

KW - Quantitative itemsets

KW - Top-k Pattern mining

UR - https://www.scopus.com/pages/publications/85125376851

U2 - 10.1007/978-3-030-95408-6_2

DO - 10.1007/978-3-030-95408-6_2

M3 - Conference contribution

AN - SCOPUS:85125376851

SN - 9783030954079

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 16

EP - 28

BT - Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings

A2 - Li, Bohan

A2 - Yue, Lin

A2 - Jiang, Jing

A2 - Chen, Weitong

A2 - Li, Xue

A2 - Long, Guodong

A2 - Fang, Fei

A2 - Yu, Han

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 2 February 2022 through 4 February 2022

ER -

Nouioua M, Fournier-Viger P, Gan W, Wu Y, Lin JCW, Nouioua F. TKQ: Top-K Quantitative High Utility Itemset Mining. In Li B, Yue L, Jiang J, Chen W, Li X, Long G, Fang F, Yu H, editors, Advanced Data Mining and Applications - 17th International Conference, ADMA 2021, Proceedings. Springer Science and Business Media Deutschland GmbH. 2022. p. 16-28. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-95408-6_2

TKQ: Top-K Quantitative High Utility Itemset Mining

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this