This paper uses the Apriori algorithm to mine association rules from user-based anime rating transactions. Support, confidence, and lift metrics are used to evaluate co-preference structures and identify strong item dependencies.