Article Preview
Top1. Introduction
Association rules have been discussed quite extensively in the data mining literature and issues related to the efficient generation of such rules from large complex dataset have been addressed. Primarily, the objective of the association rule of data mining is to discover the intrigue relationships among the items in complex, and large structured or unstructured multidimensional datasets. Generally, association rules are the data mining strategies that uncover the relationship of two entities in a dataset that assists in better learning about that data, specifically in customer buying patterns in numerous business domains. Let us consider two hypothetical examples to illustrate the concept. In a supermarket, in the entire day processing, there may be several transactions committed. Each transaction consists of the name of the items purchased. If bread, milk, and cheese, for example, together are the common items in most of the transactions, then this set {bread, milk, cheese} is termed as frequent set. So, a frequent set F can be defined as the set of items (zero or more) bought together in atleast in T transactions, a user-defined threshold. Then, it is most likely that these three items should be kept close inside the business venue, presumably, resulting in product sale increase. This concept has attained significant success in data warehouse (Data warehouse, 2013), but due to its effectiveness, is exploited in various other applications, including public health. In this paper, the use of the association rule concept is focused on its potential application to the recent public health concerns of obesity and implications of physical activity.