| Chapter 2: | Segmenting Customer Transactions Using a Pattern-Based Clustering Approach |
This is a limited free preview of this book. Please buy full access.
is convenient, it is generally not clear why it is the appropriate method for grouping customers or customer transactions. For mixture models, changing model parameters to represent the difference between segments can often oversimplify the differences and can ignore variables and patterns that are not captured by the parametric models.
In this research we study a new approach to segmenting customer transactions that is based on the idea that there may exist natural behavioral patterns in different groups of transactions. For example, a set of behavioral patterns that distinguish a group of wireless subscribers may be as follows:
The above set of three patterns may be representative of a group of consultants who travel frequently and who exhibit a set of common behavioral patterns. This example suggests that there may be natural clusters in data characterized by a set of typical behavioral patterns. In such cases, appropriate “pattern-based clustering” approaches can constitute an intuitive method for grouping customer transactions.
At the highest level, the idea is to cluster customer transactions such that patterns generated from each cluster, while similar to each other within the cluster, are very different from the patterns generated from other clusters. Customers’ behavioral patterns can have different representations. A behavioral pattern can be represented as an IF-THEN rule. It represents what a customer will do under a certain circumstance. For example, if it is the weekend, customer X will make a call longer than one hour to California using his cell phone. We can also use a collection of things (called an “itemset” in the data mining literature