Interest Clock: Time Perception in Real-Time Streaming Recommendation System
Yongchun Zhu, Jingwu Chen, Ling Chen, Yitan Li, Feng Zhang, Zuotao Liu
TL;DR
The paper addresses time-varying user preferences in real-time streaming recommender systems, where discrete time encodings can lead to instability. It introduces Interest Clock, which encodes past hour-level interests into an hour-level clock and uses Gaussian smoothing to produce a time-aware embedding $v_{clock}$ that is concatenated with standard features for prediction, formalized by $v_{clock}= \sum_{t=1}^{24} g(\delta_{time}) [v^{genre}_{t}, v^{mood}_{t}, v^{lang}_{t}]$ with $g(\delta)= \frac{1}{\sqrt{2\pi}\sigma} \exp(- (\delta-\mu)^2 /(2\sigma^2))$, $\mu=0$, $\sigma=1$. Online experiments show gains of +0.509% in Active Days and +0.758% in Duration, and offline analyses on a large industrial dataset confirm Gaussian Clock's superiority over naive/adaptive approaches, motivating deployment in production. The method demonstrates that time-aware, hour-level personalization can be effectively scaled to real-time streaming systems, improving user engagement in industry settings. Overall, Interest Clock provides a practical, universal approach to time perception in streaming recommendations with tangible business impact.
Abstract
User preferences follow a dynamic pattern over a day, e.g., at 8 am, a user might prefer to read news, while at 8 pm, they might prefer to watch movies. Time modeling aims to enable recommendation systems to perceive time changes to capture users' dynamic preferences over time, which is an important and challenging problem in recommendation systems. Especially, streaming recommendation systems in the industry, with only available samples of the current moment, present greater challenges for time modeling. There is still a lack of effective time modeling methods for streaming recommendation systems. In this paper, we propose an effective and universal method Interest Clock to perceive time information in recommendation systems. Interest Clock first encodes users' time-aware preferences into a clock (hour-level personalized features) and then uses Gaussian distribution to smooth and aggregate them into the final interest clock embedding according to the current time for the final prediction. By arming base models with Interest Clock, we conduct online A/B tests, obtaining +0.509% and +0.758% improvements on user active days and app duration respectively. Besides, the extended offline experiments show improvements as well. Interest Clock has been deployed on Douyin Music App.