The Sequence Rules algorithm finds typical sequences of item sets that frequently occur in the transactions of the input table, such as:
itemset1 >>> itemset2 >>> ... ==> itemsetN
For example,
(camera, battery) >>> (film, bag) ==> (film, battery)
means, that a customer who buys (camera, battery) on day 1, (film, bag) a week later will then buy (film, battery) next.
Items are the basic units in the rules. For example, a typical item is an individual article bought by a customer during a visit to a supermarket.
An item set is a collection of items. There are also single item sets that contain only one item.
The group ID field specifies the items that belong to a transaction. It is typically a transaction ID or timestamp.
The sequence ID field identifies the object or the entity to which the items belong. For example, on a retail scenario, this is the customer ID.