Traditional recommender systems based on Collaborative Filtering (CF) often overlook contextual information, leading to suboptimal recommendations in specific situations. Context-aware recommender Systems (CARS) aim to incorporate contextual factors such as time, location, and companions. However, existing methods face challenges related to computational complexity and selecting relevant contextual features. This paper proposes a novel context-aware recommendation approach using Energy Distance combined with Pre-Filtering Contextual Features to improve prediction accuracy while reducing computational costs. Experimental results on the MovieLens, Amazon, and Yelp datasets show significant improvements in both accuracy and efficiency compared to existing methods.