Microsoft Recommenders: Best Practices for Production-Ready Recommendation Systems

González-Fierro, Miguel; Zhang, Le

doi:10.1145/3366424.3382692

Cited by 23 publications

(11 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use industry scale recommender systems to drive the evaluation of Kairos' effectiveness. Such deep recommendation models (DRMs) have several advantages: (1) wide interest from systems research [16,60,61] (2) large customer demand and wide deployment in industry [62,63] (3) availability of public trace and artifact [17]. Table 3 lists the models and QoS as 99 𝑡ℎ tail latency target.…”

Section: Methodsmentioning

confidence: 99%

Building Heterogeneous Cloud System for Machine Learning Inference

Li¹,

Samsi²,

Gadepally³

et al. 2022

Preprint

View full text Add to dashboard Cite

Online inference is becoming a key service product for many businesses, deployed in cloud platforms to meet customer demands. Despite their revenue-generation capability, these services need to operate under tight Quality-of-Service (QoS) and cost budget constraints. This paper introduces Kairos, a novel runtime framework that maximizes the query throughput while meeting QoS target and a cost budget. Kairos designs and implements novel techniques to build a pool of heterogeneous compute hardware without online exploration overhead, and distribute inference queries optimally at runtime. Our evaluation using industry-grade deep learning (DL) models shows that Kairos yields up to 2× the throughput of an optimal homogeneous solution, and outperforms state-of-the-art schemes by up to 70%, despite advantageous implementations of the competing schemes to ignore their exploration overhead.

show abstract

Section: Methodsmentioning

confidence: 99%

Building Heterogeneous Cloud System for Machine Learning Inference

Li¹,

Samsi²,

Gadepally³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…𝑠 at timestamp 𝑡 are evolved recursively from 𝑈 (𝑡 −1) 𝑠 , affected by the last interaction 𝑌 𝑡 −1 with item 𝑉 (𝑡 −1) . • Interaction Prediction in Eqn (3). When predicting future interactions, whether long or short-term interests play a more important role depends on a wide variety of aspects, including the target item 𝑉 (𝑡 ) and the interaction history 𝒙 𝒖 of 𝑈 [47].…”

Section: User Interests Modelingmentioning

confidence: 99%

“…We implement all the models with the Microsoft Recommenders framework [3] based on TensorFlow [1]. We use the Adam optimizer [21].…”

Section: A3 Implementation Detailsmentioning

confidence: 99%

Disentangling Long and Short-Term Interests for Recommendation

Zheng,

Gao,

Chang

et al. 2022

Preprint

View full text Add to dashboard Cite

Modeling user's long-term and short-term interests is crucial for accurate recommendation. However, since there is no manually annotated label for user interests, existing approaches always follow the paradigm of entangling these two aspects, which may lead to inferior recommendation accuracy and interpretability. In this paper, to address it, we propose a Contrastive learning framework to disentangle Long and Short-term interests for Recommendation (CLSR) with self-supervision. Specifically, we first propose two separate encoders to independently capture user interests of different time scales. We then extract long-term and short-term interests proxies from the interaction sequences, which serve as pseudo labels for user interests. Then pairwise contrastive tasks are designed to supervise the similarity between interest representations and their corresponding interest proxies. Finally, since the importance of long-term and short-term interests is dynamically changing, we propose to adaptively aggregate them through an attentionbased network for prediction. We conduct experiments on two large-scale real-world datasets for e-commerce and short-video recommendation. Empirical results show that our CLSR consistently outperforms all state-of-the-art models with significant improvements: GAUC is improved by over 0.01, and NDCG is improved by over 4%. Further counterfactual evaluations demonstrate that stronger disentanglement of long and short-term interests is successfully achieved by CLSR. The code and data are available at https://github.com/tsinghua-fib-lab/CLSR.

show abstract

“…Studied inference models.To demonstrate the effectiveness and wide application range of Ribbon, the models in On the other side, MT-WND and DIEN are two representative workloads from industry personalized recommendation applications. Such applications are now consuming a large fraction of compute cycles in data centers according to the recent reports [8,10]. Therefore, studying and evaluating Ribbon's effectiveness on such models is as important as the general DNN/CNNs.…”

Section: Introductionmentioning

confidence: 99%

boringlee24/Ribbon: Ribbon: Cost-Effective and QoS-Aware Deep Learning Model Inference using a Diverse Pool of Cloud Computing Instances

2021

View full text Add to dashboard Cite

Deep learning model inference is a key service in many businesses and scientific discovery processes. This paper introduces Ribbon, a novel deep learning inference serving system that meets two competing objectives: quality-of-service (QoS) target and costeffectiveness. The key idea behind Ribbon is to intelligently employ a diverse set of cloud computing instances (heterogeneous instances) to meet the QoS target and maximize cost savings. Ribbon devises a Bayesian Optimization-driven strategy that helps users build the optimal set of heterogeneous instances for their model inference service needs on cloud computing platformsand, Ribbon demonstrates its superiority over existing approaches of inference serving systems using homogeneous instance pools. Ribbon saves up to 16% of the inference service cost for different learning models including emerging deep learning recommender system models and drug-discovery enabling models.

show abstract

Microsoft Recommenders: Best Practices for Production-Ready Recommendation Systems

Cited by 23 publications

References 1 publication

Building Heterogeneous Cloud System for Machine Learning Inference

Building Heterogeneous Cloud System for Machine Learning Inference

Disentangling Long and Short-Term Interests for Recommendation

boringlee24/Ribbon: Ribbon: Cost-Effective and QoS-Aware Deep Learning Model Inference using a Diverse Pool of Cloud Computing Instances

Contact Info

Product

Resources

About