Tenrec: A Large-scale Multipurpose Benchmark Dataset for Recommender Systems

TL;DR

Tenrec is a large-scale, multi-scenario benchmark dataset for recommender systems that includes diverse user feedback types and additional features, aiming to improve real-world applicability of RS models.

Contribution

The paper introduces Tenrec, a comprehensive and publicly available dataset with extensive user feedback and multi-scenario overlaps for advancing recommender system research.

Findings

Tenrec contains around 5 million users and 140 million interactions.

Classical baseline models perform variably across different tasks on Tenrec.

Tenrec demonstrates potential as a versatile benchmark for diverse recommendation tasks.

Abstract

Existing benchmark datasets for recommender systems (RS) either are created at a small scale or involve very limited forms of user feedback. RS models evaluated on such datasets often lack practical values for large-scale real-world applications. In this paper, we describe Tenrec, a novel and publicly available data collection for RS that records various user feedback from four different recommendation scenarios. To be specific, Tenrec has the following five characteristics: (1) it is large-scale, containing around 5 million users and 140 million interactions; (2) it has not only positive user feedback, but also true negative feedback (vs. one-class recommendation); (3) it contains overlapped users and items across four different scenarios; (4) it contains various types of user positive feedback, in forms of clicks, likes, shares, and follows, etc; (5) it contains additional features beyond the user IDs and item IDs. We verify Tenrec on ten diverse recommendation tasks by running several classical baseline models per task. Tenrec has the potential to become a useful benchmark dataset for a majority of popular recommendation tasks.