VK-LSVD: A Large-Scale Industrial Dataset for Short-Video Recommendation

2026-02-04

AI Summary

Short-video recommendation faces significant challenges in modeling rapidly evolving user interests and the scarcity of real-world, large-scale dynamic interaction data. To address these issues, this work introduces VK-LSVD, the largest publicly available short-video recommendation dataset to date, comprising over 40 billion interactions from tens of millions of users and nearly 20 million videos. VK-LSVD uniquely integrates multiple types of feedback signals, content embeddings, and contextual metadata within an open-access framework, offering an authentic representation of industrial platform dynamics. The dataset supports critical research tasks such as sequential recommendation and cold-start scenarios, and has already served as the foundation for the VK RecSys Challenge 2025, providing a high-fidelity benchmark and essential infrastructure for advancing recommender systems research.

Technology Category

• Computer Vision: Video Understanding & Activity Analysis
• Data Mining & Knowledge Management: Recommender Systems
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• User Modeling, Personalization and Recommendation: Social recommender systems and personalization for the social Web
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Short-video recommendation presents unique challenges, such as modeling rapid user interest shifts from implicit feedback, but progress is constrained by a lack of large-scale open datasets that reflect real-world platform dynamics. To bridge this gap, we introduce the VK Large Short-Video Dataset (VK-LSVD), the largest publicly available industrial dataset of its kind. VK-LSVD offers an unprecedented scale of over 40 billion interactions from 10 million users and almost 20 million videos over six months, alongside rich features including content embeddings, diverse feedback signals, and contextual metadata. Our analysis supports the dataset's quality and diversity. The dataset's immediate impact is confirmed by its central role in the live VK RecSys Challenge 2025. VK-LSVD provides a vital, open dataset to use in building realistic benchmarks to accelerate research in sequential recommendation, cold-start scenarios, and next-generation recommender systems.

Problem

• short-video recommendation
• large-scale dataset
• implicit feedback
• user interest shifts
• industrial recommender systems

Innovation

• large-scale dataset
• short-video recommendation
• implicit feedback
• industrial recommender systems
• cold-start scenarios