Multiple Utility Factors for High utility Itemset Mining

What is it about?

High Utility Itemset Mining (HUIM) was proposed in the literature to retrieve a set of items that generate high revenue for firms. For calculating the item's utility, HUIM considers both the quantity of the items and a utility factor (that represents the importance of the items) such as profit. Unlike Frequent Itemset Mining (FIM), which considers only the quantity of the items and ignores the utility factor, HUIM can generate more interesting patterns and helps firms grow their business. Most of the existing techniques for High Utility Itemset Mining consider a single factor to represent the utility of the items, this factor can be profit, cost, or any other similar factor. The itemsets retrieved by these techniques may not be of high utility when considering other factors. For example, when a dairy product such as milk is being managed over a storage counter for transactions is that the commodity's profitability may be X if sold within a few hours after being placed on the counter. If the same product remains on the cold storage counter for an entire day, the scope of managing the product on the counter includes both cooling unit costs and the issue of occupying storage space, which may preclude the addition of other items. In essence, such a condition reduces the product's profitability by causing it to generate less or no profit. To address this issue, this article proposed a new model that considers multiple factors for calculating the utility of the items which can improve the quality of the itemset recommendation.

Featured Image

Photo by Nathália Rosa on Unsplash

Photo by Nathália Rosa on Unsplash

Why is it important?

The key advantage of the proposed model is flexibility in changing the factors that could be considered for scaling the itemsets analysis, depending on the context and category of business.

Perspectives