Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{200515,
        author = {Dr S J Vivekanandan and Bujja Abigna and Dharini R and Harini N},
        title = {Novel Way to Estimate High Utility Mining using Matrix based Approach},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {2582-2592},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=200515},
        abstract = {High Utility Itemset Mining (HUIM) serves as a vital tool for extracting important data patterns from extensive transactional databases through its analysis of both item occurrence and product profitability. Traditional frequent pattern mining techniques use support values as their main measurement method while they fail to consider item utility which results in frequent pattern discovery of unimportant patterns. The current HUIM methods use unchanging threshold parameters together with multiple database scanning processes which results in increased system resource use and extended operation time. This paper presents a matrix-based method that enables effective high utility pattern extraction through the implementation of dynamic percentile-based threshold selection. The proposed method transforms transactional data into a structured utility matrix which eliminates unnecessary database scanning while enhancing system performance. The calculation of utility values utilizes quantity and profit information, whereas the calculation of support values depends on the frequency of occurrence. Percentile analysis generates dynamic thresholds which enable the system to select itemsets that hold significant meaning at any time. The proposed system classifies discovered itemsets into High Frequency High Utility (HFHU), High Frequency Low Utility (HFLU), and Low Frequency High Utility (LFHU) categories, with special emphasis on identifying low-frequency yet highly profitable patterns. The experimental results show that the proposed method enhances pattern discovery performance while it delivers valuable insights about sales patterns and profit margins which assist decision makers in their work. The system proves its enhanced capabilities through performance testing which involves large-scale transactional datasets. The interactive visualization dashboard developed through Streamlit enables users to perform intuitive analysis while FastAPI backend API layer implementation provides support for real-world applications. The matrix-based representation delivers operational efficiency advantages by decreasing the need for multiple database accesses while it enhances the system’s processing performance. The high-utility patterns which have been discovered assist organizations in making strategic decisions for their inventory control and sales planning and profit maximization activities. The proposed model improves usability through its visualization tools and API-based connectivity while enabling organizations to implement the system in their actual business operations.},
        keywords = {Utility Mining, Apriori Algorithm, High Utility Itemset Mining (HUIM), Transactional Data Analysis, Percentile-Based Thresholding, Matrix-Based Mining, Support and Utility, Low Frequency High Profit Items.},
        month = {May},
        }