Implementing a Hybrid of Efficient Algorithms For Mining Top-K High Utility Itemsets

Data mining is a methodical process of discovering data patterns and models in large data sets that involve methods at the intersection of the database system. This paper issues the popular problem of the extraction of high utility element sets (HUI) in the context of data mining. The problem of these HUIs (set of elements of high usage and value) is mainly the annoying mixture of frequent elements. Another addressable issue is the one of pattern mining which is a widespread problem in data mining, which involves searching for frequent patterns in transaction databases. Solve the problem of the set of high utility elements (HUI) requires some particular data and the state of the art of the algorithms. To store the HUI (set of high utility elements) many popular algorithms have been proposed for this problem, such as "Apriori", FP growth, etc., but now the most popular TKO algorithms (extraction of utility element sets) K in one phase) and TKU (extraction of elements sets Top-K Utility) here TKO is Top K in one phase and TKU is Top K in utility. In this paper, all the aforementioned issues have been addressed by proposing a new framework to mine k upper HUI where k is the desired number of HUI to extract. Extraction of high utility element sets is not a very common practice. Although, it is indefinitely being used in our daily lives, e.g. Online Shopping, etc. It is part of the business analysis. The main area of interest of this paper is implementing a hybrid efficient Algorithm for Top K high utility itemsets. This paper implements the hybrid of TKU and TKO with improved performance parameters overcoming the drawbacks of each algorithm