Mastering Machine Learning with Spark 2.x

Pattern mining terminology

We will start with a set of items I = {a₁, ..., a_n}, which serves as the base for all the following concepts. A transaction T is just a set of items in I, and we say that T is a transaction of length l if it contains l item. A transaction database D is a database of transaction IDs and their corresponding transactions.

To give a concrete example of this, consider the following situation. Assume that the full item set to shop from is given by I = {bread, cheese, ananas, eggs, donuts, fish, pork, milk, garlic, ice cream, lemon, oil, honey, jam, kale, salt}. Since we will look at a lot of item subsets, to make things more readable later on, we will simply abbreviate these items by their first letter, that is, we'll ...