In order to create an association rules model, the data type for all attributes in the data set must be __________.

• A. Binary
• B. Numeric
• C. Text
• D. Categorical

Answer: (A)

In association rule mining, data is treated as a collection of transactions where each possible item is either present or absent in a transaction. This binary (0/1) encoding lets the algorithm easily count how often any combination of items occurs across all transactions, which is the basis for computing support and deriving rules. If you start with numeric or text attributes, you would need preprocessing (such as discretization or one-hot encoding) to convert them into binary indicators before applying the standard approach, but the typical representation is binary.