For association rules modeling, the data type of all attributes is typically __________.

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

Answer: (B)

Association rule mining is built on transactions consisting of items, so attributes are represented as binary indicators of presence or absence. This binary encoding—each item either occurs in a transaction or not—lets algorithms efficiently count frequencies and determine frequent itemsets across many transactions. While numeric or categorical data can be used, they must be transformed into binary flags (through discretization, one-hot encoding, etc.) before standard ARM methods can operate. Text data would require even more preprocessing and isn’t the typical form used directly. Hence, the common data type for all attributes in association rules modeling is binary.