In previous blog posts (see parts 1, 2 and 3) we introduced the idea of pattern matching as a way of understanding sequences of transactions for a customer. This powerful technique can be used to both select from, and analyse a customer database and provide insight into interesting transactional behaviour.
In our previous blog posts, the cp number philippines algorithm returned a single value representing matches on the grouping table. There are cases when returning a value on the transactional table for every transaction can enable us to get a better understanding of the particular type of customer each person is.
In this blog, we illustrate some of the analytical questions that we can now seek to answer. We also revisit football analytics to answer the all-important question about which team really have the fans that have to put up with the most suffering!
In the screenshot below there are two people who both have the same number of transactions, but it is clear after quick inspection that the two people have very different holidaying behaviour. Person 1 has started with three different destinations, before settling back to going to the same destination repeatedly. Person 2 has the same initial three transactions and then switches between the three destinations.
Travel patterns
If we look for a person’s propensity to switch holiday destinations, we might look for the length of sequences of entirely different holiday destinations. We can use a set of patterns like this.