The required data type for all attributes in an association rule model in RapidMiner is _______.

• A. Binominal
• B. Numeric
• C. String
• D. Boolean

Answer: (A)

In association rule modeling, data is treated as a set of items per transaction, where each item is either present or not. This binary, presence-absence view fits cleanly when every attribute is binominal, meaning it has two possible values (for example, true/false or yes/no). With all attributes binominal, you can interpret each attribute value as an item and directly count how often different itemsets occur across transactions to compute support and confidence. This binary representation is what the RapidMiner Association Rules operator expects, allowing the algorithm to build and evaluate rules efficiently.

Choosing a different data type would require additional encoding—numeric values would need discretization, and non-binominal nominal values would require converting each category into multiple binary indicators—before the rule mining can proceed in the same straightforward way. That’s why binominal is the appropriate, directly usable form for all attributes in this context.