NEWS

We are excited to share that our paper, authored by the Rakuten Institute of Technology (RIT) in San Mateo, has been accepted for presentation at the RecSys 2024 research track!

In the field of one-class recommendation systems, training models effectively without relying on negative samples has been a significant challenge. Our work addresses this by introducing an innovative loss function that eliminates the need for negative samples, significantly speeding up training while maintaining competitive results. This advancement has the potential to streamline the development of recommendation systems, making them more efficient and scalable.

Paper Title: One-class recommendation systems with the hinge pairwise distance loss and orthogonal representations

Authors: Ramin Raziperchikolaei, Young-joo Chung

The conference will be held from October 14 to 18, 2024, in Bari, Italy.