Chao Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Chao Wang has a total of 69 co-authors. Here are the top co-authors by number of shared publications:
- Hui Xiong
The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China and The Hong Kong University of Science and Technology, Hong Kong SAR, China
16 shared publications - Yanyong Zhang
University of Science and Technology of China, Hefei, Anhui, China
6 shared publications - Dazhong Shen
Nanjing University of Aeronautics and Astronautics, Nanjing, China
5 shared publications - Chuan Qin
Computer Network Information Center, Chinese Academy of Sciences, Beijing, China and University of Chinese Academy of Sciences, Beijing, China
4 shared publications - Leilei Ding
University of Science and Technology of China
4 shared publications - Tianfu Wang
University of Science and Technology of China
4 shared publications
Latest Publications
- Open Access
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 3169-3180
GPCDR: A General Graph Prompting Framework to Adapt Single-Domain Models for Cross-Domain Recommendation
While Graph Neural Networks (GNNs) have established themselves as the state-of-the-art paradigm in recommender systems, their performance is often hampered by severe data sparsity. Cross-domain recommendation (CDR) offers a promising solution by leveraging multi-domain interactions. However, existing methods typically rely on specialized, end-to-end architectures. This isolation prevents the reuse of mature single-domain GNNs. Moreover, these methods typically employ static or uniform transfer mechanisms, failing to capture the interest heterogeneity across domains and the personalized nature of cross-domain correlations. To bridge this gap, we propose GPCDR, a general Graph Prompting framework for Cross-Domain Recommendation. Designed as a plug-and-play solution, GPCDR transforms single-domain GNN-based recommenders into effective cross-domain recommenders via the graph prompting approach. Specifically, to handle interest heterogeneity across distinct graph structures, we introduce a multi-interest graph prompting module. It employs a Mixture-of-Experts (MoE) mechanism to generate adaptive, user-specific graph prompts, enabling the backbone to capture diverse interest patterns. Furthermore, to address the rigidity of interest alignment, we design a personalized domain similarity module. Instead of enforcing uniform constraints, this module learns user-specific affinity scores to dynamically gate knowledge transfer, accommodating users with varying degrees of cross-domain consistency. Extensive experiments on four real-world datasets demonstrate that GPCDR not only significantly boosts the performance of various GNN backbones but also consistently outperforms state-of-the-art specialized CDR baselines. Our code is available at https://github.com/liujinghong123/GPCDR.
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 3169-3180
Open AccessGPCDR: A General Graph Prompting Framework to Adapt Single-Domain Models for Cross-Domain Recommendation
While Graph Neural Networks (GNNs) have established themselves as the state-of-the-art paradigm in recommender systems, their performance is often hampered by severe data sparsity. Cross-domain recommendation (CDR) offers a promising solution by leveraging multi-domain interactions. However, existing methods typically rely on specialized, end-to-end architectures. This isolation prevents the reuse of mature single-domain GNNs. Moreover, these methods typically employ static or uniform transfer mechanisms, failing to capture the interest heterogeneity across domains and the personalized nature of cross-domain correlations. To bridge this gap, we propose GPCDR, a general Graph Prompting framework for Cross-Domain Recommendation. Designed as a plug-and-play solution, GPCDR transforms single-domain GNN-based recommenders into effective cross-domain recommenders via the graph prompting approach. Specifically, to handle interest heterogeneity across distinct graph structures, we introduce a multi-interest graph prompting module. It employs a Mixture-of-Experts (MoE) mechanism to generate adaptive, user-specific graph prompts, enabling the backbone to capture diverse interest patterns. Furthermore, to address the rigidity of interest alignment, we design a personalized domain similarity module. Instead of enforcing uniform constraints, this module learns user-specific affinity scores to dynamically gate knowledge transfer, accommodating users with varying degrees of cross-domain consistency. Extensive experiments on four real-world datasets demonstrate that GPCDR not only significantly boosts the performance of various GNN backbones but also consistently outperforms state-of-the-art specialized CDR baselines. Our code is available at https://github.com/liujinghong123/GPCDR.
- Open Access
- 29 October 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 18  (1) , 1-27
Against Political Polarization: A Unified Framework for Tracing Evolving Political Ideologies on Social Media
The rapid growth of social media has greatly influenced political discourse, highlighting the need to understand individual political ideologies and their temporal dynamics. This task faces challenges such as data scarcity, abundant non-political content, costly and bias-prone manual annotation, and difficulty in modeling future ideological inclinations. To address these issues, we propose TSN4PI , a unified framework for tracking the evolution of political ideologies on social media. It includes two core modules. The PIDN uses large language models with style transfer and unsupervised domain adaptation to enable robust ideology detection and filter irrelevant content from noisy, cross-domain data. The PIPN employs temporal graph neural networks to predict future ideological shifts, enabling comprehensive analysis of ideology presence, intensity, and evolution. We release two large-scale datasets for noncommercial research use to facilitate further work. Extensive case studies on multiple platforms (X and Truth Social) validate the effectiveness of TSN4PI and provide empirical insights into political polarization and the evolution of online ideologies. Our findings offer a nuanced perspective, advancing both methodological development and empirical understanding in this field.
- 29 October 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 18  (1) , 1-27
Open AccessAgainst Political Polarization: A Unified Framework for Tracing Evolving Political Ideologies on Social Media
The rapid growth of social media has greatly influenced political discourse, highlighting the need to understand individual political ideologies and their temporal dynamics. This task faces challenges such as data scarcity, abundant non-political content, costly and bias-prone manual annotation, and difficulty in modeling future ideological inclinations. To address these issues, we propose TSN4PI , a unified framework for tracking the evolution of political ideologies on social media. It includes two core modules. The PIDN uses large language models with style transfer and unsupervised domain adaptation to enable robust ideology detection and filter irrelevant content from noisy, cross-domain data. The PIPN employs temporal graph neural networks to predict future ideological shifts, enabling comprehensive analysis of ideology presence, intensity, and evolution. We release two large-scale datasets for noncommercial research use to facilitate further work. Extensive case studies on multiple platforms (X and Truth Social) validate the effectiveness of TSN4PI and provide empirical insights into political polarization and the evolution of online ideologies. Our findings offer a nuanced perspective, advancing both methodological development and empirical understanding in this field.
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 9801-9812
MobilityBench: A Benchmark for Evaluating Route-Planning Agents in Real-World Mobility Scenarios
Route-planning agents powered by large language models (LLMs) have emerged as a promising paradigm for supporting everyday human mobility through natural language interaction and tool-mediated decision making. However, systematic evaluation in real-world mobility settings is hindered by diverse routing demands, non-deterministic mapping services, and limited reproducibility. In this study, we introduce MobilityBench, a scalable benchmark for evaluating LLM-based route-planning agents in real-world mobility scenarios. MobilityBench is constructed from large-scale, anonymized real user queries collected from Amap and covers a broad spectrum of route-planning intents across multiple cities worldwide. To enable reproducible, end-to-end evaluation, we design a deterministic API-replay sandbox that eliminates environmental variance from live services. We further propose a multi-dimensional evaluation protocol centered on outcome validity, complemented by assessments of instruction understanding, planning, tool use, and efficiency. Using MobilityBench, we evaluate multiple LLM-based route-planning agents across diverse real-world mobility scenarios and provide an in-depth analysis of their behaviors and performance. Our findings reveal that current models perform competently on Basic information retrieval and Route Planning tasks, yet struggle considerably with Preference-Constrained Route Planning, underscoring significant room for improvement in personalized mobility applications. We publicly release the benchmark data, evaluation toolkit, and documentation at https://github.com/AMAP-ML/MobilityBench.
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 9801-9812
Open AccessMobilityBench: A Benchmark for Evaluating Route-Planning Agents in Real-World Mobility Scenarios
Route-planning agents powered by large language models (LLMs) have emerged as a promising paradigm for supporting everyday human mobility through natural language interaction and tool-mediated decision making. However, systematic evaluation in real-world mobility settings is hindered by diverse routing demands, non-deterministic mapping services, and limited reproducibility. In this study, we introduce MobilityBench, a scalable benchmark for evaluating LLM-based route-planning agents in real-world mobility scenarios. MobilityBench is constructed from large-scale, anonymized real user queries collected from Amap and covers a broad spectrum of route-planning intents across multiple cities worldwide. To enable reproducible, end-to-end evaluation, we design a deterministic API-replay sandbox that eliminates environmental variance from live services. We further propose a multi-dimensional evaluation protocol centered on outcome validity, complemented by assessments of instruction understanding, planning, tool use, and efficiency. Using MobilityBench, we evaluate multiple LLM-based route-planning agents across diverse real-world mobility scenarios and provide an in-depth analysis of their behaviors and performance. Our findings reveal that current models perform competently on Basic information retrieval and Route Planning tasks, yet struggle considerably with Preference-Constrained Route Planning, underscoring significant room for improvement in personalized mobility applications. We publicly release the benchmark data, evaluation toolkit, and documentation at https://github.com/AMAP-ML/MobilityBench.
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 6080-6091
CoGen: A Two-Stage Unified Multi-modal Framework for Bundle Construction via Discrete Semantics Transfer
Bundle construction aims to assemble sets of mutually compatible items that collectively satisfy user preferences. This problem naturally encompasses two interconnected tasks: bundle completion, which captures intra-bundle compatibility, and bundle generation, which constructs personalized bundles. However, existing approaches typically treat these tasks in isolation, neglecting their inherent synergy: completion models learn compatibility in entangled continuous spaces that are difficult to transfer, while generation models often suffer from sparse supervision and lack explicit compatibility constraints, leading to semantically incoherent bundles. To bridge this gap, we propose CoGen, a two-stage unified multi-modal framework for bundle construction. The core idea of CoGen is to transform compatibility-aware continuous representations learned from completion into discrete semantics to guide bundle generation. Specifically, CoGen first learns disentangled multi-modal item representations and performs bundle completion to capture intra-bundle compatibility. The learned representations are then discretized via residual quantization (RQ) to obtain explicit discrete semantics that make compatibility patterns reusable for generation. To further improve bundle generation quality, CoGen adopts a semantic-anchored non-autoregressive generator to generate bundles based on the learned discrete semantics. Extensive experiments on four public multi-modal datasets verify the effectiveness of CoGen compared to several state-of-the-art approaches. Our code is available at https://github.com/y2q109/CoGen.
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 6080-6091
Open AccessCoGen: A Two-Stage Unified Multi-modal Framework for Bundle Construction via Discrete Semantics Transfer
Bundle construction aims to assemble sets of mutually compatible items that collectively satisfy user preferences. This problem naturally encompasses two interconnected tasks: bundle completion, which captures intra-bundle compatibility, and bundle generation, which constructs personalized bundles. However, existing approaches typically treat these tasks in isolation, neglecting their inherent synergy: completion models learn compatibility in entangled continuous spaces that are difficult to transfer, while generation models often suffer from sparse supervision and lack explicit compatibility constraints, leading to semantically incoherent bundles. To bridge this gap, we propose CoGen, a two-stage unified multi-modal framework for bundle construction. The core idea of CoGen is to transform compatibility-aware continuous representations learned from completion into discrete semantics to guide bundle generation. Specifically, CoGen first learns disentangled multi-modal item representations and performs bundle completion to capture intra-bundle compatibility. The learned representations are then discretized via residual quantization (RQ) to obtain explicit discrete semantics that make compatibility patterns reusable for generation. To further improve bundle generation quality, CoGen adopts a semantic-anchored non-autoregressive generator to generate bundles based on the learned discrete semantics. Extensive experiments on four public multi-modal datasets verify the effectiveness of CoGen compared to several state-of-the-art approaches. Our code is available at https://github.com/y2q109/CoGen.
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 7797-7808
CEDAR: Controlled and Event-Driven Demand Forecasting via Residual Decomposition
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 7797-7808
Open AccessCEDAR: Controlled and Event-Driven Demand Forecasting via Residual Decomposition
- Open Access
- 20 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 879-890
Rethinking Popularity Bias in Collaborative Filtering via Analytical Vector Decomposition
Popularity bias fundamentally undermines the personalization capabilities of collaborative filtering (CF) models, causing them to disproportionately recommend popular items while neglecting users' genuine preferences for niche content. While existing approaches treat this as an external confounding factor, we reveal that popularity bias is an intrinsic geometric artifact of Bayesian Pairwise Ranking (BPR) optimization in CF models. Through rigorous mathematical analysis, we prove that BPR systematically organizes item embeddings along a dominant ''popularity direction'' where embedding magnitudes directly correlate with interaction frequency. This geometric distortion forces user embeddings to simultaneously handle two conflicting tasks—expressing genuine preference and calibrating against global popularity—trapping them in suboptimal configurations that favor popular items regardless of individual tastes. We propose Directional Decomposition and Correction (DDC), a universally applicable framework that surgically corrects this embedding geometry through asymmetric directional updates. DDC guides positive interactions along personalized preference directions while steering negative interactions away from the global popularity direction, disentangling preference from popularity at the geometric source. Extensive experiments across multiple BPR-based architectures demonstrate that DDC significantly outperforms state-of-the-art debiasing methods, reducing training loss to less than 5% of heavily-tuned baselines while achieving superior recommendation quality and fairness. Code is available in https://github.com/LingFeng-Liu-AI/DDC.
- 20 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 879-890
Open AccessRethinking Popularity Bias in Collaborative Filtering via Analytical Vector Decomposition
Popularity bias fundamentally undermines the personalization capabilities of collaborative filtering (CF) models, causing them to disproportionately recommend popular items while neglecting users' genuine preferences for niche content. While existing approaches treat this as an external confounding factor, we reveal that popularity bias is an intrinsic geometric artifact of Bayesian Pairwise Ranking (BPR) optimization in CF models. Through rigorous mathematical analysis, we prove that BPR systematically organizes item embeddings along a dominant ''popularity direction'' where embedding magnitudes directly correlate with interaction frequency. This geometric distortion forces user embeddings to simultaneously handle two conflicting tasks—expressing genuine preference and calibrating against global popularity—trapping them in suboptimal configurations that favor popular items regardless of individual tastes. We propose Directional Decomposition and Correction (DDC), a universally applicable framework that surgically corrects this embedding geometry through asymmetric directional updates. DDC guides positive interactions along personalized preference directions while steering negative interactions away from the global popularity direction, disentangling preference from popularity at the geometric source. Extensive experiments across multiple BPR-based architectures demonstrate that DDC significantly outperforms state-of-the-art debiasing methods, reducing training loss to less than 5% of heavily-tuned baselines while achieving superior recommendation quality and fairness. Code is available in https://github.com/LingFeng-Liu-AI/DDC.
- Open Access
- 20 April 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Knowledge Discovery From Data
- Vol. 20  (4) , 1-31
Graph-based Prompt Learning with Mixture of Experts for Multi-task Corporate Profiling
Corporate profiling serves as a critical analytical tool for modern enterprises, enabling data-driven decision-making in investment strategies, risk assessment, and strategic planning. It requires integrating quantitative metrics, qualitative insights, and network relationships to capture a company’s role in the business ecosystem. However, traditional methods struggle to synthesize heterogeneous data and model complex interdependencies among corporations, news, and market dynamics, often addressing these aspects in isolation. To address these challenges, this article introduces Financial Graph-based Mixture of Experts Prompt Learning (FGMPL), an innovative framework that unifies graph prompt learning with a multi-task paradigm for corporate profile modeling. The proposed framework reformulates node- and edge-level tasks into a coherent graph-level representation and employs multi-view contrastive learning to effectively integrate textual details with relational structures. Moreover, a novel Financial Multi-Experts Prompting mechanism—with learnable tokens coupled with a Mixture of Experts (MoE) design—is presented to enhance the processing of heterogeneous graph data and bridge the gap between pre-training and downstream tasks. To further improve adaptability, a meta-learning-based prompt tuning strategy is incorporated, enabling rapid transition to various downstream applications. Extensive experiments on real-world financial graphs show that FGMPL consistently outperforms strong pre-training and graph-prompting baselines across corporate performance prediction, relationship prediction, and news classification in both full-data and few-shot settings. In addition, cross-market transfer on a NASDAQ dataset and interpretability/efficiency analyses further demonstrate its robustness and practical applicability.
- 20 April 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Knowledge Discovery From Data
- Vol. 20  (4) , 1-31
Open AccessGraph-based Prompt Learning with Mixture of Experts for Multi-task Corporate Profiling
Corporate profiling serves as a critical analytical tool for modern enterprises, enabling data-driven decision-making in investment strategies, risk assessment, and strategic planning. It requires integrating quantitative metrics, qualitative insights, and network relationships to capture a company’s role in the business ecosystem. However, traditional methods struggle to synthesize heterogeneous data and model complex interdependencies among corporations, news, and market dynamics, often addressing these aspects in isolation. To address these challenges, this article introduces Financial Graph-based Mixture of Experts Prompt Learning (FGMPL), an innovative framework that unifies graph prompt learning with a multi-task paradigm for corporate profile modeling. The proposed framework reformulates node- and edge-level tasks into a coherent graph-level representation and employs multi-view contrastive learning to effectively integrate textual details with relational structures. Moreover, a novel Financial Multi-Experts Prompting mechanism—with learnable tokens coupled with a Mixture of Experts (MoE) design—is presented to enhance the processing of heterogeneous graph data and bridge the gap between pre-training and downstream tasks. To further improve adaptability, a meta-learning-based prompt tuning strategy is incorporated, enabling rapid transition to various downstream applications. Extensive experiments on real-world financial graphs show that FGMPL consistently outperforms strong pre-training and graph-prompting baselines across corporate performance prediction, relationship prediction, and news classification in both full-data and few-shot settings. In addition, cross-market transfer on a NASDAQ dataset and interpretability/efficiency analyses further demonstrate its robustness and practical applicability.
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 6481-6492
MCLMR: A Model-Agnostic Causal Learning Framework for Multi-Behavior Recommendation
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 6481-6492
Open AccessMCLMR: A Model-Agnostic Causal Learning Framework for Multi-Behavior Recommendation
- Open Access
- 3 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 3216-3227
Structure-Enhanced Protein Instruction Tuning: Towards General-Purpose Protein Understanding with LLMs
- 3 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 3216-3227
Open AccessStructure-Enhanced Protein Instruction Tuning: Towards General-Purpose Protein Understanding with LLMs
- Open Access
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 570-580
Can LLMs Enhance Fairness in Recommendation Systems? A Data Augmentation Approach
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 570-580
Open AccessCan LLMs Enhance Fairness in Recommendation Systems? A Data Augmentation Approach