Abstract

Current recommendation systems face fundamental challenges in processing complex user preferences and managing large-scale data, particularly in restaurant recommendations where traditional methods struggle with context integration and computational efficiency. Our study presents a quantum-enhanced context-aware recommendation system that combines Ordering Points To Identify Cluster Structure (OPTICS) clustering, Long Short Term Memory (LSTM), Recurrent Neural Network (RNN), Quantum Machine Learning, and Whale Optimization to address these limitations. Applied to the NYC Restaurant Dataset, our framework utilizes quantum-enhanced clustering with advanced word embeddings, quantum-augmented Long Short Term Memory Recurrent Neural Network (LSTM RNN) for temporal pattern analysis, and quantum similarity metrics for restaurant matching. The system achieves remarkable performance metrics: 89.54% accuracy, Root Mean Square Error (RMSE) of 0.2876, Mean Absolute Error (MAE) of 0.2234, precision of 0.9021, and recall of 0.8876, with an <em>F</em>-value of 189.6754 and a statistically significant <em>P</em>-value of 3.45e–32. These results demonstrate substantial improvements over classical methods in processing speed, feature representation, and similarity calculations, establishing a new benchmark in restaurant recommendation systems.