Search by tag «Recommendation systems» 3 results

• Russian Scientists Propose More Accurate Update Method for Recommendation Systems, No Retraining Needed

  Researchers at AI VK and ITMO’s Laboratory for Multiagent Modeling and Adaptive Intelligence have proposed a more precise way to update recommendation algorithms that will help reduce the computational costs of updating ML models while preserving backward compatibility.

  23.04.2026

• Russian Scientists Help Recommendation Systems Understand Users Better

  Researchers from ITMO University, Sberbank’s Practical AI Center, AIRI Institute, and Innopolis University have found a way to distill deep semantic knowledge from large language models (LLMs) into lightweight recommendation systems. For users, this means more accurate recommendations for products, movies, and other content and a stable service performance, with recommendations being generated as quickly as with the original compact model. The findings of the study were reported at ECIR 2026.

  20.04.2026

• MTS to Teach ITMO Students to Develop Recommendation Systems in Python

  MTS, a Russian telecom company, has launched a course on the development of recommendation systems using Python for the joint online Master’s program Machine Learning Engineering delivered by ITMO University and Napoleon IT. Over three weeks, big data specialists from MTS will give lectures, seminars, and Q&A sessions, as well as offer tasks based on the company’s actual cases using open-source data and frameworks.

  17.11.2022