Data mining uses techniques to sift through massive amounts of data to tell us something informative. In rule induction, if-then "rules" are generated based on patterns found. This is often applied to store receipts, for instance. In this example, we see two rules being generated, if a customer buys bread then the customer will buy peanut butter (this rule has a high frequency and high utility) and if a customer buys wine then the customer will buy brown rice (this rule has a very high frequency but a low utility, so is perhaps not as useful). When it comes to rule induction, we would expect to use thousands or even millions of receipts, not the 8 receipts shown here.