A team of students from the Politecnico di Milano, coordinated by Maurizio Ferrari Dacrema, researcher at the Department of Electronics, Information and Bioengineering, won the prestigious RecSys Challenge 2023 as the best academic team. The team is composed by Paolo Basso, Arturo Benedetti, Nicola Cecere, Alessandro Maranelli, Salvatore Marragony, Samuele Peri, Andrea Riboni, Alessandro Verosimile, Davide Zanutto.
The RecSys Challenge 2023 lasted for three months, with its final phase in Singapore from September 18-22, 2023 as part of the 17th ACM Conference on Recommender System, and was organized by the Indian social media platform ShareChat in collaboration with the Indian Institute of Management Visakhapatnam, Huawei and Amazon.
This year's challenge focused on online advertising, the optimization of thesteps through which a user subjected to an advertisement turns into a customer by deciding to make a purchase, as well as to the protection of user privacy. Tackling this problem requires to develop a complex personalization process, which opens several challenges related to the modeling of the user behavior, to the use of large-scale machine learning as well as safeguarding user privacy.
The participating teams were provided with a real-world dataset from the ShareChat and Moj apps, generated from about 10 million random users who visited the apps over a three-month period.
A total of 500 teams participated in the competitions, including some large companies such as NVIDIA and Intel as well as universities such as the Tsinghua University. The team from Politecnico di Milano has described its proposed solution in a paper titled “Pessimistic Rescaling and Distribution Shift of Boosting Models for Impression-Aware Online Advertising Recommendation”.
