Are Convolutional Sequential Recommender Systems Still Competitive? Introducing New Models and Insights

Federico Siciliano, Antonio Purificato, Filippo Betello, Nicola Tonellotto, Fabrizio Silvestri

Abstract: Sequential Recommender Systems (SRSs) have predominantly shifted to neural-based models. Despite significant advances, Convolutional Neural Network (CNN)-based SRSs are rapidly being overshadowed by the more performant attention-based models. In this paper, we present a novel modification of two widely used CNN-based SRSs, Caser and CosRec. We improve their training by adapting the convolution and pooling operations so that they can be trained simultaneously on the whole input sequence rather than just on the last element. Our experimental results show that these modified CNN-based models achieve up to +65% in NDCG@10 compared to their original versions. Furthermore, they are also more competitive with SASRec, one of the most used attention-based SRS, surpassing it by up to +53% in NDCG@10. These findings suggest that CNN-based SRSs, with appropriate modifications, warrant further investigation and may offer viable alternatives to current attention-based approaches. Our code is available at https://github.com/antoniopurificato/recsys_conv_conf.

External IDs:dblp:conf/ijcnn/SicilianoPBTS25