Article
Computer Science, Artificial Intelligence
Hai Duong, Tien Hoang, Thong Tran, Tin Truong, Bac Le, Philippe Fournier-Viger
Summary: Closed high utility itemsets (CHUIs) and maximal high utility itemsets (MaxHUIs) are important concise representations of HUIs. Mining these representations is crucial for generating meaningful high utility association rules. However, existing algorithms suffer from long runtimes, high memory usage, and scalability issues. To address this, this paper proposes two efficient algorithms that can mine these representations faster.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Automation & Control Systems
Thong Tran, Hai Duong, Tin Truong, Bac Le
Summary: The discovery of frequent closed high utility itemsets (FCHUIs) and frequent generators of high utility itemsets (FGHUIs) is important for providing essential summaries and generating nonredundant high utility association rules. This paper proposes a novel approach using a new weak lower bound (wlbu) to efficiently mine these itemsets and presents two new algorithms that outperform existing algorithms.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Trinh D. D. Nguyen, Loan T. T. Nguyen, Lung Vu, Bay Vo, Witold Pedrycz
Summary: This research addresses the problem of mining high-utility itemsets in dynamic unit profit databases, introducing a novel algorithm iEFIM-Closed that outperforms state-of-the-art algorithms in sparse databases.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Multidisciplinary Sciences
Damla Oguz
Summary: High-utility itemset mining aims to discover sets of items that are sold together with utility values that exceed a minimum threshold. The method considers internal and external utility values of the itemsets, and their symmetric effects in determining high-utility itemsets. A proposed asymmetric approach focuses on high external utility values while ignoring internal utility values, leading to more efficient discovery and highlighting the fundamental impact of external utility values.
Article
Engineering, Multidisciplinary
M. S. Bhuvaneswari, N. Balaganesh, K. Muneeswaran
Summary: HAUI mining improves upon HUI mining by using average utility to find itemsets more efficiently, resulting in faster processing time and reduced memory usage.
SADHANA-ACADEMY PROCEEDINGS IN ENGINEERING SCIENCES
(2022)
Article
Engineering, Electrical & Electronic
M. S. Bhuvaneswari, N. Balaganesh, K. Muneeswaran
Summary: High Average Utility Itemset mining addresses the limitations of HUIM by taking itemset length into account, improving utility estimation accuracy, enhancing processing efficiency with pruning algorithms, and reducing processing time through a multi-threaded parallel approach.
IETE JOURNAL OF RESEARCH
(2022)
Article
Computer Science, Artificial Intelligence
Jun-Feng Qu, Philippe Fournier-Viger, Mengchi Liu, Bo Hang, Chunyang Hu
Summary: This paper proposes a new algorithm called Hamm for mining high utility itemsets. Hamm utilizes a TV (prefix Tree and utility Vector) structure to mine high utility itemsets in a one-phase manner without candidate generation. Experimental results show that Hamm outperforms other algorithms in terms of performance.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Information Systems
Krishan Kumar Sethi, Dharavath Ramesh, Munesh Chandra Trivedi
Summary: HUI mining is a data mining technique to discover profitable patterns, and this research proposes new strategies and a distributed algorithm to make it suitable for big data processing. Experimental results demonstrate that the proposed algorithm outperforms existing algorithms.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Yoonji Baek, Unil Yun, Heonho Kim, Jongseong Kim, Bay Vo, Tin Truong, Zhi-Hong Deng
Summary: A new technique for high utility pattern mining considering data noises is proposed in this paper, which calculates trustworthy ranges for patterns using a utility tolerance factor to extract robust high utility patterns from noisy databases. Experimental results demonstrate that the proposed algorithm outperforms competitors in terms of runtime, memory usage, and scalability.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Information Systems
Mohammed A. Fouad, Wedad Hussein, Sherine Rady, Philip S. Yu, Tarek F. Gharib
Summary: This paper introduces a new problem of mining reliable high utility patterns and proposes an efficient approach called RUPM to solve it. Experimental results show that compared to traditional algorithms, the RUPM approach can obtain patterns with higher reliability and provide pruning strategies to reduce runtime and memory usage.
Article
Computer Science, Artificial Intelligence
Wei Fang, Qiang Zhang, Hengyang Lu, Jerry Chun-Wei Lin
Summary: This study proposes an improved binary particle swarm optimization (HUIM-IBPSO) for high-utility itemset mining (HUIM), addressing the issues of exponential growth search space and time-consuming process in traditional exact algorithms. The proposed approach incorporates multiple adjustment strategies to keep the same HUIs, enhance search ability, avoid premature convergence, and improve efficiency in mining HUIs.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Loan T. T. Nguyen, Thang Mai, Giao-Huy Pham, Unil Yun, Bay Vo
Summary: Many researchers are exploring and using a new trend in data mining called high occupancy itemset mining. This research applies an occupancy measure to a support-based mining framework, bringing benefits to decision support systems and enabling managers to visualize reports and analyze data more efficiently.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Peng Wu, Xinzheng Niu, Philippe Fournier-Viger, Cheng Huang, Bing Wang
Summary: High utility itemset mining is a key problem in data mining, aiming to find itemsets with high importance or profit in a database for decision-making support. This paper proposes to improve the efficiency of utility-list construction through a set of bitwise operations and a new data structure, leading to faster mining of high utility itemsets.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Hardware & Architecture
Lili Chen, Wensheng Gan, Qi Lin, Shuqiang Huang, Chien-Ming Chen
Summary: Mobile edge computing has brought new opportunities and challenges to data science, and "OHUQI" is an efficient method for discovering high-utility quantitative itemsets, reducing unnecessary scans and designing pruning strategies to shrink the search space.
JOURNAL OF SUPERCOMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Zaihe Cheng, Wei Fang, Wei Shen, Jerry Chun-Wei Lin, Bo Yuan
Summary: High-utility itemset mining is an important task in data mining for retrieving meaningful patterns. Existing algorithms suffer from storage and time overheads. To address this, we propose an efficient algorithm based on simplified utility-list structure, which effectively reduces the number of candidates, memory usage, and execution time by introducing techniques like simplified utility-list, repeated pruning, and extension utility.
APPLIED INTELLIGENCE
(2023)