With growing digital globalization and online purchases, the need for effective recommendation systems is increasing to enable consumers to find appropriate products. Current e-commerce websites list searched products, but a better solution is suggesting alternative products with similarities, enhancing user experience and enhancing the chances of purchase by 55%. The present work suggests a product recommendation system based on the Amazon clothing dataset that holds 330,000 apparel products. Natural Language Processing (NLP) examines product names, and Compatible Finite Element-Interpolated Neural Networks (CFEINNs) anticipate similar products. Trainable Fractional Fourier Transform (T-FrFT) is used to extract features from product images, and Transfer learning is used to improve model performance. Based on the calculation of distances between feature vectors, the system generates product recommendations with the most similar proximity. Leveraging Randomized Smoothing (LRS) strengthens data preprocessing and resilience even more, providing precise and appropriate recommendations.