Cong Fu
Cong Fu has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
- Anxiang Zeng
Nanyang technological university, Singapore, Singapore
8 shared publications - Yabo Ni
Nanyang technological university, Singapore, Singapore
8 shared publications - Kun Wang
Shopee Pte. Ltd., Beijing, China
4 shared publications - Yuxuan Zhu
Shopee Pte. Ltd., Shanghai, China
4 shared publications - Jiahua Wu
Shopee Pte. Ltd., Singapore, Singapore
3 shared publications - Kangle Wu
Shopee Pte. Ltd., Singapore, Singapore
3 shared publications
Latest Publications
- Open Access
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 8818-8826
ELBO TDS : Temporal Distribution Shift Generalization for Industry-Scale Recommender Systems
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 8818-8826
Open AccessELBO TDS : Temporal Distribution Shift Generalization for Industry-Scale Recommender Systems
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 7117-7128
OnePiece: Bringing Context Engineering and Reasoning to Industrial Cascade Ranking System
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 7117-7128
Open AccessOnePiece: Bringing Context Engineering and Reasoning to Industrial Cascade Ranking System
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 8053-8064
OneRank: Unified Transformer-Native Ranking Architecture for Multi-Task Recommendation
Multi-task learning (MTL) is essential in recommender systems to enable complementary learning among diverse user feedback. While modern industrial practices have shifted from DNNs to Transformer-centric architectures to strengthen sequence modeling and scaling capacity, they still decouple feature encoding from multi-task prediction, treating the Transformer as a task-agnostic encoder. This design fundamentally limits the performance and scalability by (1) creating an information bottleneck under heterogeneous task objectives, (2) inducing gradient interference that leads to the seesaw phenomenon, and (3) forcing a dataflow transition in which attention-based, context-adaptive representation learning is converted to static feed-forward task prediction with incompatible information read–write dynamics. In this paper, we propose OneRank, a Transformer-native multi-task ranking framework that eliminates the encoder–predictor separation and introduces task-private channels for both forward representation learning and backward optimization, enabling task-specialized learning while minimizing inter-task interferences. In the forward pass, OneRank learns task-specific representations in a bottom-up manner through task-conditioned information selection, candidate–aware contextualization, and controlled cross-task interaction. In the backward pass, cross-task gradient detachment isolates task-private parameter updates from shared knowledge extraction modules, preventing negative transfer. Finally, we replace static task-specific MLP scorers with a dynamic matching-based scoring formulation for context-aware personalized ranking. By internalizing multi-task reasoning pathways within the Transformer stack, OneRank establishes a new architectural paradigm with unified modeling and scalable computation design. Extensive offline and online experiments on large-scale industrial datasets demonstrate that OneRank significantly outperforms state-of-the-art baselines across multiple tasks, with substantial improvements in ranking effectiveness while maintaining computational efficiency.
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 8053-8064
Open AccessOneRank: Unified Transformer-Native Ranking Architecture for Multi-Task Recommendation
Multi-task learning (MTL) is essential in recommender systems to enable complementary learning among diverse user feedback. While modern industrial practices have shifted from DNNs to Transformer-centric architectures to strengthen sequence modeling and scaling capacity, they still decouple feature encoding from multi-task prediction, treating the Transformer as a task-agnostic encoder. This design fundamentally limits the performance and scalability by (1) creating an information bottleneck under heterogeneous task objectives, (2) inducing gradient interference that leads to the seesaw phenomenon, and (3) forcing a dataflow transition in which attention-based, context-adaptive representation learning is converted to static feed-forward task prediction with incompatible information read–write dynamics. In this paper, we propose OneRank, a Transformer-native multi-task ranking framework that eliminates the encoder–predictor separation and introduces task-private channels for both forward representation learning and backward optimization, enabling task-specialized learning while minimizing inter-task interferences. In the forward pass, OneRank learns task-specific representations in a bottom-up manner through task-conditioned information selection, candidate–aware contextualization, and controlled cross-task interaction. In the backward pass, cross-task gradient detachment isolates task-private parameter updates from shared knowledge extraction modules, preventing negative transfer. Finally, we replace static task-specific MLP scorers with a dynamic matching-based scoring formulation for context-aware personalized ranking. By internalizing multi-task reasoning pathways within the Transformer stack, OneRank establishes a new architectural paradigm with unified modeling and scalable computation design. Extensive offline and online experiments on large-scale industrial datasets demonstrate that OneRank significantly outperforms state-of-the-art baselines across multiple tasks, with substantial improvements in ranking effectiveness while maintaining computational efficiency.
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 5998-6009
ManCAR: Manifold-Constrained Latent Reasoning with Adaptive Test-Time Computation for Sequential Recommendation
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 5998-6009
Open AccessManCAR: Manifold-Constrained Latent Reasoning with Adaptive Test-Time Computation for Sequential Recommendation
- Open Access
- 2 April 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Management of Data
- Vol. 4 (1) , 1-25
Reveal Hidden Pitfalls and Navigate Next Generation of Vector Similarity Search from Task-Centric Views: [Experiments & Analysis]
Vector Similarity Search (VSS) in high-dimensional spaces is rapidly emerging as core functionality in next-generation database systems for numerous data-intensive services -- from embedding lookups in large language models (LLMs), to semantic information retrieval and recommendation engines. Current benchmarks, however, evaluate VSS primarily on the recall-latency trade-off against a ground truth defined solely by distance metrics, neglecting how retrieval quality ultimately impacts downstream tasks. This disconnect can mislead both academic research and industrial practice. We present Iceberg, a holistic benchmark suite for end-to-end evaluation of VSS methods in realistic application contexts. From a task-centric view, Iceberg uncovers the Information Loss Funnel , which identifies three principal sources of end-to-end performance degradation: (1) Embedding Loss during feature extraction; (2) Metric Misuse , where distances poorly reflect task relevance; (3) Data Distribution Sensitivity , highlighting index robustness across skews and modalities. For a more comprehensive assessment, Iceberg spans eight diverse datasets across key domains such as image classification, face recognition, text retrieval, and recommendation systems. Each dataset, ranging from 1M to 100M vectors, includes rich, task-specific labels and evaluation metrics, enabling assessment of retrieval algorithms within the full application pipeline rather than in isolation. Iceberg benchmarks 13 state-of-the-art VSS methods and re-ranks them based on application-level metrics, revealing substantial deviations from traditional rankings derived purely from recall-latency evaluations. Building on these insights, we define a set of task-centric meta-features and derive an interpretable decision tree to guide practitioners in selecting and tuning VSS methods for their specific workloads. Finally, we outline promising avenues for future research aimed at closing the gap between synthetic VSS benchmarks and production deployments.
- 2 April 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Management of Data
- Vol. 4 (1) , 1-25
Open AccessReveal Hidden Pitfalls and Navigate Next Generation of Vector Similarity Search from Task-Centric Views: [Experiments & Analysis]
Vector Similarity Search (VSS) in high-dimensional spaces is rapidly emerging as core functionality in next-generation database systems for numerous data-intensive services -- from embedding lookups in large language models (LLMs), to semantic information retrieval and recommendation engines. Current benchmarks, however, evaluate VSS primarily on the recall-latency trade-off against a ground truth defined solely by distance metrics, neglecting how retrieval quality ultimately impacts downstream tasks. This disconnect can mislead both academic research and industrial practice. We present Iceberg, a holistic benchmark suite for end-to-end evaluation of VSS methods in realistic application contexts. From a task-centric view, Iceberg uncovers the Information Loss Funnel , which identifies three principal sources of end-to-end performance degradation: (1) Embedding Loss during feature extraction; (2) Metric Misuse , where distances poorly reflect task relevance; (3) Data Distribution Sensitivity , highlighting index robustness across skews and modalities. For a more comprehensive assessment, Iceberg spans eight diverse datasets across key domains such as image classification, face recognition, text retrieval, and recommendation systems. Each dataset, ranging from 1M to 100M vectors, includes rich, task-specific labels and evaluation metrics, enabling assessment of retrieval algorithms within the full application pipeline rather than in isolation. Iceberg benchmarks 13 state-of-the-art VSS methods and re-ranks them based on application-level metrics, revealing substantial deviations from traditional rankings derived purely from recall-latency evaluations. Building on these insights, we define a set of task-centric meta-features and derive an interpretable decision tree to guide practitioners in selecting and tuning VSS methods for their specific workloads. Finally, we outline promising avenues for future research aimed at closing the gap between synthetic VSS benchmarks and production deployments.
- Open Access
- 3 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 3921-3932
Embed Progressive Implicit Preference in Unified Space for Deep Collaborative Filtering
- 3 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 3921-3932
Open AccessEmbed Progressive Implicit Preference in Unified Space for Deep Collaborative Filtering
- Open Access
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 2341-2350
Stitching Inner Product and Euclidean Metrics for Topology-aware Maximum Inner Product Search
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 2341-2350
Open AccessStitching Inner Product and Euclidean Metrics for Topology-aware Maximum Inner Product Search
- Open Access
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4974-4985
Residual Multi-Task Learner for Applied Ranking
Modern e-commerce platforms rely heavily on modeling diverse user feedback to provide personalized services. Consequently, multi-task learning has become an integral part of their ranking systems. However, existing multi-task learning methods encounter two main challenges: some lack explicit modeling of task relationships, resulting in inferior performance, while others have limited applicability due to being computationally intensive, having scalability issues, or relying on strong assumptions. To address these limitations and better fit our real-world scenario, pre-rank in Shopee Search, we introduce in this paper ResFlow, a lightweight multi-task learning framework that enables efficient cross-task information sharing via residual connections between corresponding layers of task networks. Extensive experiments on datasets from various scenarios and modalities demonstrate its superior performance and adaptability over state-of-the-art methods. The online A/B tests in Shopee Search showcase its practical value in large-scale industrial applications, evidenced by a 1.29% increase in OPU (order-per-user) without additional system latency. ResFlow is now fully deployed in the pre-rank module of Shopee Search. To facilitate efficient online deployment, we propose a novel offline metric Weighted Recall@K, which aligns well with our online metric OPU, addressing the longstanding online-offline metric misalignment issue. Besides, we propose to fuse scores from the multiple tasks additively when ranking items, which outperforms traditional multiplicative fusion.
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 4974-4985
Open AccessResidual Multi-Task Learner for Applied Ranking
Modern e-commerce platforms rely heavily on modeling diverse user feedback to provide personalized services. Consequently, multi-task learning has become an integral part of their ranking systems. However, existing multi-task learning methods encounter two main challenges: some lack explicit modeling of task relationships, resulting in inferior performance, while others have limited applicability due to being computationally intensive, having scalability issues, or relying on strong assumptions. To address these limitations and better fit our real-world scenario, pre-rank in Shopee Search, we introduce in this paper ResFlow, a lightweight multi-task learning framework that enables efficient cross-task information sharing via residual connections between corresponding layers of task networks. Extensive experiments on datasets from various scenarios and modalities demonstrate its superior performance and adaptability over state-of-the-art methods. The online A/B tests in Shopee Search showcase its practical value in large-scale industrial applications, evidenced by a 1.29% increase in OPU (order-per-user) without additional system latency. ResFlow is now fully deployed in the pre-rank module of Shopee Search. To facilitate efficient online deployment, we propose a novel offline metric Weighted Recall@K, which aligns well with our online metric OPU, addressing the longstanding online-offline metric misalignment issue. Besides, we propose to fuse scores from the multiple tasks additively when ranking items, which outperforms traditional multiplicative fusion.
- Top Cited Papers
- 23 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 44 (01628828) , 1
High Dimensional Similarity Search with Satellite System Graph: Efficiency, Scalability, and Unindexed Query Compatibility
Approximate nearest neighbor search (ANNS) in high-dimensional space is essential in database and information retrieval. Recently, there has been a surge of interest in exploring efficient graph-based indices for the ANNS problem. Among them, navigating spreading-out graph (NSG) provides fine theoretical analysis and achieves state-of-the-art performance. However, we find there are several limitations with NSG: 1) NSG has no theoretical guarantee on nearest neighbor search when the query is not indexed in the database; and 2) NSG is too sparse which harms the search performance. In addition, NSG suffers from high indexing complexity. To address above problems, we propose the satellite system graphs (SSG) and a practical variant NSSG. Specifically, we propose a novel pruning strategy to produce SSGs from the complete graph. SSGs define a new family of MSNETs in which the out-edges of each node are distributed evenly in all directions. Each node in the graph builds effective connections to its neighborhood omnidirectionally, whereupon we derive SSG's excellent theoretical properties for both indexed and unindexed queries. We can adaptively adjust the sparsity of an SSG with a hyper-parameter to optimize the search performance. Further, NSSG is proposed to reduce the indexing complexity of the SSG for large-scale applications. Both theoretical and extensive experimental analysis are provided to demonstrate the strengths of the proposed approach over the existing representative algorithms. Our code has been released at https://github.com/ZJULearning/SSG .
- 23 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 44 (01628828) , 1
Top Cited PapersHigh Dimensional Similarity Search with Satellite System Graph: Efficiency, Scalability, and Unindexed Query Compatibility
Approximate nearest neighbor search (ANNS) in high-dimensional space is essential in database and information retrieval. Recently, there has been a surge of interest in exploring efficient graph-based indices for the ANNS problem. Among them, navigating spreading-out graph (NSG) provides fine theoretical analysis and achieves state-of-the-art performance. However, we find there are several limitations with NSG: 1) NSG has no theoretical guarantee on nearest neighbor search when the query is not indexed in the database; and 2) NSG is too sparse which harms the search performance. In addition, NSG suffers from high indexing complexity. To address above problems, we propose the satellite system graphs (SSG) and a practical variant NSSG. Specifically, we propose a novel pruning strategy to produce SSGs from the complete graph. SSGs define a new family of MSNETs in which the out-edges of each node are distributed evenly in all directions. Each node in the graph builds effective connections to its neighborhood omnidirectionally, whereupon we derive SSG's excellent theoretical properties for both indexed and unindexed queries. We can adaptively adjust the sparsity of an SSG with a hyper-parameter to optimize the search performance. Further, NSSG is proposed to reduce the indexing complexity of the SSG for large-scale applications. Both theoretical and extensive experimental analysis are provided to demonstrate the strengths of the proposed approach over the existing representative algorithms. Our code has been released at https://github.com/ZJULearning/SSG .