Two-Tower Neural Network for Personalizing Service Recommendation in Cloud Environment

Effectively recommending cloud services requires a deep understanding of both user requirements and the wide range of available cloud offerings. This paper introduces a novel two-tower neural network architecture to tackle this challenge. By leveraging the power of neural networks, we distill intricate user and cloud service information into concise vector representations, enabling the refinement of highly personalized recommendations. Our proposed two-tower architecture is specifically designed for efficient scalability within the demanding context of large-scale cloud environments. Each tower, dedicated to users and cloud services respectively, integrates a comprehensive set of relevant features and interaction data to generate dense vector embeddings. Subsequently, cloud services are ranked based on the calculated similarity between their embeddings and those of the respective users. The proposed two-tower model was evaluated on the WSdream dataset for both regression and classification tasks. Experimental results consistently demonstrated superior performance compared to state-of-the-art recommendation techniques, including k-nearest neighbors and matrix factorization.