This study explores a blended learning model for piano music education, merging traditional classroom instruction with an online personalized learning system to boost resource efficiency, student engagement, and learning outcomes. Utilizing the item collaborative filtering (CF) and learning style filtering recommendation algorithms, we tailored teaching materials to individual student needs, significantly improving match accuracy between resources and learners. Results from implementing this optimized hybrid teaching approach showed a 15% increase in course ratings, a 20% rise in student participation, and a marked enhancement in learners’ interest in piano studies. Additionally, learner satisfaction soared by 25% due to the learning style-based algorithm’s ability to personalize resource allocation. This research underscores the effectiveness of combining blended teaching models with personalized learning systems in elevating piano education quality and efficacy.