Optimized Multi-Stage Access Pattern Discovery Using a Hybrid Weighted FP-Growth Algorithm

Author

Mr. M.Rajkumar, Dr. J.M.Dhayashankar

Abstract

 The rapid growth of multi-stage access platforms has generated complex user interaction logs that require efficient and accurate pattern-mining techniques. Traditional Apriori-based algorithms perform poorly on such data due to repeated database scans, excessive candidate itemset generation, and an inability to incorporate the varying importance of user-interaction stages. To address these challenges, this study introduces a Hybrid Weighted FP-Growth with Stage-Aware Pattern Mining (HW-FP-SAPM) model and evaluates it on a ally generated benchmark dataset consisting of 1,050 records across 350 users and three interaction stages. Each stage captures distinct item access behaviors, enabling weighted analysis of early, mid, and final user actions. The proposed model employs dynamic stage-based weighting, compressed FP-Tree construction, and conditional pattern mining to extract high-value frequent itemsets with significantly reduced computational cost. Experimental analysis on the dataset shows that the HW-FP-SAPM approach eliminates candidate explosion, reduces scan overhead to two passes, and improves frequent-pattern accuracy by capturing low-frequency but high-importance itemsets within later user stages. The results demonstrate that stage-aware weighting not only enhances association-rule precision but also provides deeper insights into evolving user behavior across access stages. This work contributes a scalable, efficient, and domain-adaptive framework for multi-stage behavioral mining, offering strong potential for applications in e-commerce, digital learning platforms, healthcare monitoring systems, and other interactive multi-session environments.

Keywords

Multi-Stage Pattern Mining, Weighted FP-Growth, Stage-Aware Analysis, Sequential User Behavior, Association Rule Mining

Full Text:

Download Paper PDF

References

1.Han J., Pei J. & Yin Y., “Mining Frequent Patterns without Candidate
Generation,” in Proceedings of the 2000 ACM SIGMOD International Conference on Management of Data, 2000, pp. 1–12.
2.Liu T., “The New Algorithms of Weighted Association Rules based on Apriori and FP-Growth Methods,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 12, no. 5, 2014, pp. 4071–4078.
3.Danessh M. S., Balasubramanian C. & Duraiswamy K., “Similarity Data Item Set Approach: An Encoded Temporal Data Base Technique,” arXiv preprint arXiv:1003.4076, 2010.
4.“A Survey on Frequent Pattern Mining Methods: Apriori, Eclat, FP-Growth,” A. Saxena & S. Gadhiya, IJEDR, vol. 2, no. 1, 2014.
5.Banumathi A. & Dhurga Devi M., “Comparative Study of Frequent Itemset Mining Algorithms Apriori and FP-Growth,” International Journal of Development Research, 2018.
6.Priadharsni A. & Ramaraj E., “Comparative Study of Revised FP–Growth, Weighted Apriori and Fuzzy Apriori Algorithm,” International Journal of Engineering Trends and Technology (IJETT), vol. 13, no. 6, 2014, pp. 261–265.
7.“Mining frequent itemsets/association rules from E-Commerce transactional dataset: A Comparative Analysis of Candidate Generation (Apriori) vs Pattern Growth
(FP-Growth),” Applied Sciences, 2023.
8.H. Li, J. Han & J. Pei, “An improved frequent pattern growth method for mining association rules,” Expert Systems with Applications, vol. 38, issue 5, 2011, pp. 5154–5161.
9.Zixian Wu & Gang Fang, “The Research on the Improvement of FP-growth Algorithm,” Applied Information Technology & Research, 2022.

10.Wang C. S., Chang J. Y. & colleagues, “MISFP-growth: Hadoop-Based Frequent Pattern Mining with Multiple Item Support,” Applied Sciences, 2019.
11.Smith J. M., “An Efficient Parallel FP-Growth Algorithm for Big Data Association Rule
Mining,” Journal of Computer Science and Software Applications, vol. 4, no. 1, 2024, pp. 1–8.
12.Senthilkumar A. & D. Hari Prasad, “An efficient FP-Growth based association rule mining algorithm using Hadoop MapReduce,” Indian Journal of Science and Technology, vol. 13, issue 34, 2020, pp. 3561–3571.
13.Ait-Mlouk A., Agouti T. & Gharnati F., “Mining and Prioritization of Association Rules for Big Data: Multi-Criteria Decision Analysis Approach,” Journal of Big Data, vol. 4, article 42, 2017.
14.Akter A. & Hasan R., “Frequent pattern mining with improved Apriori and FP-Growth algorithms for big data applications,” Journal of Data Mining and Analysis, vol. 1, issue 1, 2024.
15.Shabtay L., Yaari R. & Dattner I., “A Guided FP-growth Algorithm for Multitude-Targeted Mining of Big Data,” arXiv preprint arXiv:1803.06632, 2018.
16.Gan W., Lin J. C.-W., Fournier-Viger P., Chao H.-C. & Yu P. S., “A Survey of Parallel Sequential Pattern Mining,” arXiv preprint arXiv:1805.10515, 2018.
17.Dizon F. S. V., Farinas S. K. R., Mahinay R. J. T. H. Jr. et al., “Learning of High Dengue Incidence with Clustering and FP-Growth Algorithm using WHO Historical Data,” arXiv preprint arXiv:1901.11376, 2019.
18.Wu S., “An association rule-based approach for frequent item mining of multi-stage access data,” Discover Computing, vol. 28, article 139, 2025.
19.(PaWI team), “PaWI — A Framework for Weighted Itemset Mining on Big Data,” Politecnico di Torino Research Report, 2024.
20.Cai T. & colleagues, “Weighted Based Frequent and Infrequent Pattern Mining Model for Real-time E-Commerce Databases,” Advances in Modelling and Analysis B, Vol. 62, No. 2, 2023.

Share your valuable work from Social Media Buttons