A Combinatorial Approach for High Utility Item Set Mining using FRUP and Direct Discovery Approach without Candidate Generation

Mansi Jaiswal, Vijay Prakash

Abstract: The main purpose of data mining and analytics is to find novel, potentially useful patterns that can be utilized in real-world applications to derive beneficial knowledge. For identifying and evaluating the usefulness of different kinds of patterns, many techniques/constraints have been proposed, such as support, confidence, sequence order, and utility parameters (e. g. , weight, price, profit, quantity, etc. ). In recent years, there has been an increasing demand for utility-oriented pattern mining (UPM). UPM is a vital task, with numerous high-impact applications, including cross-marketing, e-commerce, finance, medical, and biomedical applications. In this research work we have undertook two different approach as proposed in [1] and [2]. One approach uses RUP/FRUP growth algorithm while the other method uses direct discovery algorithm which does not uses candidate generation. The FRUP/FRUP approach is more extensive in a sense that not only it is helpful in determining frequent itemset but it also helps in finding the utility of the item set in a more cohesive manner. We used Matlab programming environment to combine the two approaches. The experimental results show that RUP/FRUP when combined with direct discovery approach gives better results.

Keywords: RUP/FRUP-GROWTH algorithm, HUI, data mining, apriori, big data