Highway Service Areas Recommendation System with Collaborative Filtering using Channel Attention Module based Graph Convolutional Neural Network

Collaborative Filtering (CF) is the most effective approach utilized through highway service areas recommendation system to address the information overload problem. Highway service areas recommendation provides an efficient solution to extract important data from large amount of data produced through broad cardinality of user devices. However, the rating matrix is widely sparse and it significantly affects the performance of the recommendation system. Hence, this research proposes the Channel Attention Module based Graph Convolutional Neural Network (CAM-GCNN) approach for the highway service areas recommendation model. Initially, the three public datasets such as Amazon AToys, Amazon AMusic and Goodbooks-10k are utilized to validate the effectiveness of the proposed method. Then, Stacked Denoising Autoencoder (SDAE) approach is utilized to extract the latent of user and item features. The experimental results shows that the proposed CAM-GCNN attains better results such as precision of 0.934,0.945 and 0.957 on Amazon AToys, Amazon AMusic and Goodbooks-10k datasets respectively. These results are superior than the existing methods like CF with Neural Network (NN) and Hierarchical Reinforcement Learning (HRL)