Blending Sequential Embeddings, Graphs, and Engineered Features: 4th Place Solution in RecSys Challenge 2025

TL;DR

This paper presents a high-performing solution for RecSys Challenge 2025 that combines sequential user embeddings, graph neural networks, feature interactions, and engineered features to improve behavioral modeling across multiple tasks.

Contribution

It introduces an integrated approach combining sequential, graph, and feature engineering techniques for universal behavioral modeling in recommendation systems.

Findings

Achieved 4th place in RecSys Challenge 2025

Demonstrated effectiveness of combining multiple modeling techniques

Improved user embedding quality across diverse tasks

Abstract

This paper describes the 4th-place solution by team ambitious for the RecSys Challenge 2025, organized by Synerise and ACM RecSys, which focused on universal behavioral modeling. The challenge objective was to generate user embeddings effective across six diverse downstream tasks. Our solution integrates (1) a sequential encoder to capture the temporal evolution of user interests, (2) a graph neural network to enhance generalization, (3) a deep cross network to model high-order feature interactions, and (4) performance-critical feature engineering.