International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 03 | Mar 2025
p-ISSN: 2395-0072
www.irjet.net
Enhancement Of Apriori Algorithm Using Hash Based Technique And Transaction Reduction T. Haripreethi 1, G.V.S. Ananthanath 2 1Student, Department of MCA, KMMIPS, Tirupati
2 Associative Professor, Department. of Computer Science, KMMIPS, Tirupati
---------------------------------------------------------------------***--------------------------------------------------------------------This study presents an enhanced algorithm, Apriori2, Abstract - The continuous expansion of digital data has
which integrates hash-based pruning and transaction reduction to improve the efficiency of association rule mining. [5] By reducing the number of database scans and optimizing candidate generation, Apriori2 achieves faster execution times while maintaining accuracy.[6] A case study on crime analysis demonstrates the algorithm’s practical applicability in uncovering crime patterns with minimal computational cost.
necessitated advanced techniques for uncovering meaningful patterns to support decision-making. Association rule mining, a key component of data mining, plays a vital role in extracting valuable insights from large datasets. The Apriori algorithm is widely recognized for its effectiveness in identifying frequent patterns within transactional databases. However, its efficiency is hindered by repeated database scans, leading to increased computational complexity. This paper introduces an optimized version of the Apriori algorithm, termed Apriori2, which integrates hash-based techniques and transaction reduction to enhance performance. A case study on crime analysis demonstrates the algorithm’s capability to extract significant patterns while reducing computational overhead. The findings confirm that Apriori2 is a scalable and efficient solution for real-world applications across multiple domains.
The remainder of this paper is structured as follows: Section 2 provides an overview of association rule mining and its applications. Section 3 reviews the limitations of the traditional Apriori algorithm and its existing enhancements. Section 4 introduces the Apriori2 algorithm, detailing its methodology and implementation. Section 5 presents an experimental evaluation comparing Apriori2 with conventional approaches. Finally, Section 6 concludes the study by summarizing key findings and discussing future research directions.
Key Words: Data mining, Association rule mining, Apriori algorithm, Frequent item sets, Crime analysis, Transaction reduction, Hash-based techniques.
2.Literature Survey
1.INTRODUCTION
Association rule mining has been a key area of research in data mining, enabling the discovery of frequent patterns and relationships within large datasets. Various algorithms have been proposed to improve the efficiency of frequent itemset mining, particularly in addressing the computational limitations of the traditional Apriori algorithm.
The rapid digital transformation has led to an enormous surge in data generation across industries. To harness the potential of this data, sophisticated mining techniques are required to extract useful insights. Data mining, a crucial step in the knowledge discovery process, enables organizations to analyze large datasets and uncover hidden correlations. One of the essential tasks in data mining is association rule mining, which helps in identifying frequently occurring patterns among data items.[1]
2.1. Traditional Approaches and Apriori Algorithm The Apriori algorithm, proposed by Agrawal and Srikant (1994), is a fundamental method for ARM, employing a breadth-first search strategy to generate frequent itemsets iteratively. However, Apriori suffers from significant computational inefficiencies, including:
Association rule mining plays a fundamental role in predictive analytics, allowing businesses and institutions to make data-driven decisions.[2] The Apriori algorithm, introduced by Agrawal and Srikant, is a pioneering method in this field. It systematically generates frequent itemsets from transactional databases, providing valuable rules for various applications.[3] However, the conventional Apriori algorithm suffers from performance bottlenecks due to its multiple database scans. Several enhancements have been proposed to overcome these limitations, including hashbased techniques and transaction reduction strategies.[4]
© 2025, IRJET
|
Impact Factor value: 8.315
•
High candidate generation cost – leading to exponential growth in large datasets.
•
Multiple database scans – causing performance degradation with increasing dataset size.
To address these inefficiencies, several improvements and alternative algorithms have been proposed.
|
ISO 9001:2008 Certified Journal
|
Page 912