Which data mining technique is specifically used to discover relationships among items without an explicit target variable?

• A. Machine learning
• B. Association rule learning
• C. Classification
• D. Clustering

Answer: (B)

Association rule learning focuses on uncovering relationships between items that co-occur in data, especially in transactional data, without a predefined target variable. It works by finding frequent itemsets and turning them into if-then rules (for example, if a customer buys bread and butter, they may also buy jam), using metrics like support, confidence, and lift to measure significance. This makes it ideal for understanding how items are related rather than predicting a single outcome. In contrast, machine learning is a broad field that includes methods that may require a target variable; classification explicitly predicts a label for each instance, and clustering groups similar instances without a predefined target but does not focus on item-item relationships in the same explicit way. So the technique that best fits discovering relationships among items without an explicit target is association rule learning.