Effective algorithm for frequent pattern mining

• Published in: 2017 International Conference on IoT and Application (ICIOT) pp. 1 - 5
• Main Authors: Aditya, S. P., Hemanth, M., Lakshmikanth, C. K., Suneetha, K. R.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-05-2017
• Subjects: Algorithm design and analysis; Association Rule; Data mining; Data preprocessing; Support Count; Web Mining; Web pages; Web Usage Mining
• DOI: 10.1109/ICIOTA.2017.8073615

Apriori algorithm is an influential data mining algorithm which generates a list of most frequent web pages visited. Due to fast changing contents of database one needs an algorithm which is real time. The major drawback of Apriori algorithm is that, it needs to scan main database each and every time to generate frequent patterns which results in more usage of memory and execution time, hence in order to reduce execution time and usage of memory a lot of research has been conducted to improve Apriori Algorithm. Towards improving Apriori, a modified version is proposed in this paper to generate frequent patterns. This enables finding patterns rather than going back to the database at every pass. This limits the number of scan and also the number of total combinations is brought down from 2 N to 2 N-2 . This considerably reduces memory usage as well as execution speed and makes real time pattern discovery possible.