Jing Dong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jing Dong has a total of 38 co-authors. Here are the top co-authors by number of shared publications:
Ziyu Shen
Fu Foundation School of Engineering and Applied Science, Columbia University, New York, USA
2 shared publications- Baoli Lu
Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China
1 shared publication - Chao Yao
Ira A. Fulton Schools of Engineering, Arizona State University, Tempe, AZ 85281, USA
1 shared publication - Chengrui Zhou
Fu Foundation School of Engineering and Applied Science Columbia University New York NY USA
1 shared publication - Chengrui Zhou
Fu Foundation School of Engineering and Applied Science, Columbia University, New York, USA
1 shared publication - Dan Li1 shared publication
Latest Publications
- 21 July 2026
- Published by Elsevier in Information Fusion
- Vol. 137
Robust multi-view pedestrian detection via density-guided occlusion augmentation and adaptive feature fusion
- 21 July 2026
- Published by Elsevier in Information Fusion
- Vol. 137
Robust multi-view pedestrian detection via density-guided occlusion augmentation and adaptive feature fusion
- Open Access
- 7 April 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Graph attention and text semantics improve personalized recommendation
With the increasing challenge of information overload, personalized recommendation systems play an essential role in delivering relevant content to users. Traditional collaborative filtering methods often suffer from data sparsity and cold-start problems, which limit their effectiveness in real-world applications. To address these issues, this paper proposes a personalized recommendation model that integrates graph attention mechanisms with auxiliary textual information. User–item interactions are modeled on a knowledge graph, where graph attention networks are employed to capture multi-hop user interests, while graph convolution is used to aggregate item-side neighborhood information. Textual data associated with users and items are encoded into semantic embeddings and incorporated to enrich the initial representations of graph entities. Experiments conducted on the Book-Crossing and MovieLens-1M datasets demonstrate that the proposed model achieves superior performance in terms of AUC, F1-score, and Top-K recall compared with several state-of-the-art baselines. The results indicate that combining graph-based modeling with textual semantic enhancement can effectively improve recommendation accuracy and robustness under sparse data conditions.
- 7 April 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessGraph attention and text semantics improve personalized recommendation
With the increasing challenge of information overload, personalized recommendation systems play an essential role in delivering relevant content to users. Traditional collaborative filtering methods often suffer from data sparsity and cold-start problems, which limit their effectiveness in real-world applications. To address these issues, this paper proposes a personalized recommendation model that integrates graph attention mechanisms with auxiliary textual information. User–item interactions are modeled on a knowledge graph, where graph attention networks are employed to capture multi-hop user interests, while graph convolution is used to aggregate item-side neighborhood information. Textual data associated with users and items are encoded into semantic embeddings and incorporated to enrich the initial representations of graph entities. Experiments conducted on the Book-Crossing and MovieLens-1M datasets demonstrate that the proposed model achieves superior performance in terms of AUC, F1-score, and Top-K recall compared with several state-of-the-art baselines. The results indicate that combining graph-based modeling with textual semantic enhancement can effectively improve recommendation accuracy and robustness under sparse data conditions.
- Open Access
- 2 April 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (4) , e0345461
TFP-Net: A temporal-feature-prototypical network for CRM optimization and cold-start mitigation
With the continuous growth of user behavior data on e-commerce platforms, effectively predicting user behavior, providing personalized product recommendations, and addressing the cold-start problem have become key challenges in recommendation systems. To address these issues, this paper proposes a novel customer relationship management model, TFP-Net, which integrates Temporal Graph Attention Mechanism (TGAT), Deep Feature Interaction Module (DeepFM), and Prototypical Network (ProtoNet) to enhance the performance of recommendation systems. Specifically, TFP-Net uses TGAT to capture the temporal features of user behavior, handling the dynamic changes in user actions across different time periods. The DeepFM module learns the complex non-linear relationships between users and products, while the ProtoNet optimizes recommendations in cold-start scenarios, mitigating the data sparsity issues for new users and products. In the experiments, we evaluate the model on the Taobao User Behavior Dataset and the Amazon Product Dataset. The results demonstrate that TFP-Net outperforms traditional baseline models on both datasets, especially in cold-start scenarios, with a performance improvement of 1.5% to 2.1% over the best existing models, namely DGN-JBP on the Taobao dataset and GACE on the Amazon Product Dataset. TFP-Net achieves an accuracy of 87.8% on the Taobao dataset and 88.1% on the Amazon dataset. Additionally, the model demonstrates superior computational efficiency, with lower training time and inference latency compared to other models, proving its potential in large-scale data processing. TFP-Net effectively addresses user behavior prediction and product recommendation on e-commerce platforms, excelling in cold-start scenarios and computational efficiency. It provides a new solution for personalized recommendations and customer relationship management.
- 2 April 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (4) , e0345461
Open AccessTFP-Net: A temporal-feature-prototypical network for CRM optimization and cold-start mitigation
With the continuous growth of user behavior data on e-commerce platforms, effectively predicting user behavior, providing personalized product recommendations, and addressing the cold-start problem have become key challenges in recommendation systems. To address these issues, this paper proposes a novel customer relationship management model, TFP-Net, which integrates Temporal Graph Attention Mechanism (TGAT), Deep Feature Interaction Module (DeepFM), and Prototypical Network (ProtoNet) to enhance the performance of recommendation systems. Specifically, TFP-Net uses TGAT to capture the temporal features of user behavior, handling the dynamic changes in user actions across different time periods. The DeepFM module learns the complex non-linear relationships between users and products, while the ProtoNet optimizes recommendations in cold-start scenarios, mitigating the data sparsity issues for new users and products. In the experiments, we evaluate the model on the Taobao User Behavior Dataset and the Amazon Product Dataset. The results demonstrate that TFP-Net outperforms traditional baseline models on both datasets, especially in cold-start scenarios, with a performance improvement of 1.5% to 2.1% over the best existing models, namely DGN-JBP on the Taobao dataset and GACE on the Amazon Product Dataset. TFP-Net achieves an accuracy of 87.8% on the Taobao dataset and 88.1% on the Amazon dataset. Additionally, the model demonstrates superior computational efficiency, with lower training time and inference latency compared to other models, proving its potential in large-scale data processing. TFP-Net effectively addresses user behavior prediction and product recommendation on e-commerce platforms, excelling in cold-start scenarios and computational efficiency. It provides a new solution for personalized recommendations and customer relationship management.
- Open Access
- 1 January 2026
- Published by IGI Global Scientific Publishing in Journal of Organizational and End User Computing
- Vol. 38 (1) , 1-30
A Deep Learning-Based Model for User Emotion Recognition and Behavior Guidance Design
EMoGuideNet is a deep-learning model that integrates user emotion and behavior in multi-turn interactions, overcoming the current limitation of treating them as separate processes. By simultaneously streaming emotional conditions and behavioral responses, it captures their interactive dynamics. The model processes multimodal inputs (textual/acoustic) through a recurrent emotion tracking mechanism, enabling temporal emotion modeling, and uses multi-task learning to predict behavior. Evaluated on four benchmarks (MELD, IEMOCAP, DailyDialog, ReDial), it outperforms CoCoLM, SER-Meta, and BERT4Rec, achieving a statistically significant 3.9% improvement in emotion-recognition accuracy and 4.6% in behavioral Hit@10 on MELD (p < 0.01). The model also demonstrates robustness to noise and generalizes well to unseen settings, highlighting the effectiveness of emotion-conditioned behavior modeling. These advancements enhance adaptive dialogue systems by enabling responses with greater emotional relevance and situational awareness.
- 1 January 2026
- Published by IGI Global Scientific Publishing in Journal of Organizational and End User Computing
- Vol. 38 (1) , 1-30
Open AccessA Deep Learning-Based Model for User Emotion Recognition and Behavior Guidance Design
EMoGuideNet is a deep-learning model that integrates user emotion and behavior in multi-turn interactions, overcoming the current limitation of treating them as separate processes. By simultaneously streaming emotional conditions and behavioral responses, it captures their interactive dynamics. The model processes multimodal inputs (textual/acoustic) through a recurrent emotion tracking mechanism, enabling temporal emotion modeling, and uses multi-task learning to predict behavior. Evaluated on four benchmarks (MELD, IEMOCAP, DailyDialog, ReDial), it outperforms CoCoLM, SER-Meta, and BERT4Rec, achieving a statistically significant 3.9% improvement in emotion-recognition accuracy and 4.6% in behavioral Hit@10 on MELD (p < 0.01). The model also demonstrates robustness to noise and generalizes well to unseen settings, highlighting the effectiveness of emotion-conditioned behavior modeling. These advancements enhance adaptive dialogue systems by enabling responses with greater emotional relevance and situational awareness.
- Top Cited PapersOpen Access
From Black Box to Glass Box: A Practical Review of Explainable Artificial Intelligence (XAI)
Explainable Artificial Intelligence (XAI) has become essential as machine learning systems are deployed in high-stakes domains such as security, finance, and healthcare. Traditional models often act as “black boxes”, limiting trust and accountability. Traditional models often act as “black boxes”, limiting trust and accountability. However, most existing reviews treat explainability either as a technical problem or a philosophical issue, without connecting interpretability techniques to their real-world implications for security, privacy, and governance. This review fills that gap by integrating theoretical foundations with practical applications and societal perspectives. define transparency and interpretability as core concepts and introduce new economics-inspired notions of marginal transparency and marginal interpretability to highlight diminishing returns in disclosure and explanation. Methodologically, we examine model-agnostic approaches such as LIME and SHAP, alongside model-specific methods including decision trees and interpretable neural networks. We also address ante-hoc vs. post hoc strategies, local vs. global explanations, and emerging privacy-preserving techniques. To contextualize XAI’s growth, we integrate capital investment and publication trends, showing that research momentum has remained resilient despite market fluctuations. Finally, we propose a roadmap for 2025–2030, emphasizing evaluation standards, adaptive explanations, integration with Zero Trust architectures, and the development of self-explaining agents supported by global standards. By combining technical insights with societal implications, this article provides both a scholarly contribution and a practical reference for advancing trustworthy AI.
Top Cited PapersOpen AccessFrom Black Box to Glass Box: A Practical Review of Explainable Artificial Intelligence (XAI)
Explainable Artificial Intelligence (XAI) has become essential as machine learning systems are deployed in high-stakes domains such as security, finance, and healthcare. Traditional models often act as “black boxes”, limiting trust and accountability. Traditional models often act as “black boxes”, limiting trust and accountability. However, most existing reviews treat explainability either as a technical problem or a philosophical issue, without connecting interpretability techniques to their real-world implications for security, privacy, and governance. This review fills that gap by integrating theoretical foundations with practical applications and societal perspectives. define transparency and interpretability as core concepts and introduce new economics-inspired notions of marginal transparency and marginal interpretability to highlight diminishing returns in disclosure and explanation. Methodologically, we examine model-agnostic approaches such as LIME and SHAP, alongside model-specific methods including decision trees and interpretable neural networks. We also address ante-hoc vs. post hoc strategies, local vs. global explanations, and emerging privacy-preserving techniques. To contextualize XAI’s growth, we integrate capital investment and publication trends, showing that research momentum has remained resilient despite market fluctuations. Finally, we propose a roadmap for 2025–2030, emphasizing evaluation standards, adaptive explanations, integration with Zero Trust architectures, and the development of self-explaining agents supported by global standards. By combining technical insights with societal implications, this article provides both a scholarly contribution and a practical reference for advancing trustworthy AI.
- Open Access
LLM-Augmented Symptom Analysis for Cardiovascular Disease Risk Prediction: A Clinical NLP
Timely identification and accurate risk stratification of cardiovascular disease (CVD) remain essential for reducing global mortality. While existing prediction models primarily leverage structured data, unstructured clinical notes contain valuable early indicators. This study introduces a novel LLM-augmented clinical NLP pipeline that employs domain-adapted large language models for symptom extraction, contextual reasoning, and correlation from free-text reports. Our approach integrates cardiovascular-specific fine-tuning, prompt-based inference, and entity-aware reasoning. Evaluations on MIMIC-III and CARDIO-NLP datasets demonstrate improved performance in precision, recall, F1-score, and AUROC, with high clinical relevance (kappa = 0.82) assessed by cardiologists. Challenges such as contextual hallucination, which occurs when plausible information contracts with provided source, and temporal ambiguity, which is related with models struggling with chronological ordering of events are addressed using prompt engineering and hybrid rule-based verification. This work underscores the potential of LLMs in clinical decision support systems (CDSS), advancing early warning systems and enhancing the translation of patient narratives into actionable risk assessments.
Open AccessLLM-Augmented Symptom Analysis for Cardiovascular Disease Risk Prediction: A Clinical NLP
Timely identification and accurate risk stratification of cardiovascular disease (CVD) remain essential for reducing global mortality. While existing prediction models primarily leverage structured data, unstructured clinical notes contain valuable early indicators. This study introduces a novel LLM-augmented clinical NLP pipeline that employs domain-adapted large language models for symptom extraction, contextual reasoning, and correlation from free-text reports. Our approach integrates cardiovascular-specific fine-tuning, prompt-based inference, and entity-aware reasoning. Evaluations on MIMIC-III and CARDIO-NLP datasets demonstrate improved performance in precision, recall, F1-score, and AUROC, with high clinical relevance (kappa = 0.82) assessed by cardiologists. Challenges such as contextual hallucination, which occurs when plausible information contracts with provided source, and temporal ambiguity, which is related with models struggling with chronological ordering of events are addressed using prompt engineering and hybrid rule-based verification. This work underscores the potential of LLMs in clinical decision support systems (CDSS), advancing early warning systems and enhancing the translation of patient narratives into actionable risk assessments.
- Open Access
- 13 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 53-59
Research on Personalized Financial Product Recommendation by Integrating Large Language Models and Graph Neural Networks
- 13 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 53-59
Open AccessResearch on Personalized Financial Product Recommendation by Integrating Large Language Models and Graph Neural Networks
- Open Access
Research on Personalized Financial Product Recommendation by Integrating Large Language Models and Graph Neural Networks
With the rapid growth of fintech, personalized financial product recommendations have become increasingly important. Traditional methods like collaborative filtering or content-based models often fail to capture users' latent preferences and complex relationships. We propose a hybrid framework integrating large language models (LLMs) and graph neural networks (GNNs). A pre-trained LLM encodes text data (e.g., user reviews) into rich feature vectors, while a heterogeneous user-product graph models interactions and social ties. Through a tailored message-passing mechanism, text and graph information are fused within the GNN to jointly optimize embeddings. Experiments on public and real-world financial datasets show our model outperforms standalone LLM or GNN in accuracy, recall, and NDCG, with strong interpretability. This work offers new insights for personalized financial recommendations and cross-modal fusion in broader recommendation tasks.
Open AccessResearch on Personalized Financial Product Recommendation by Integrating Large Language Models and Graph Neural Networks
With the rapid growth of fintech, personalized financial product recommendations have become increasingly important. Traditional methods like collaborative filtering or content-based models often fail to capture users' latent preferences and complex relationships. We propose a hybrid framework integrating large language models (LLMs) and graph neural networks (GNNs). A pre-trained LLM encodes text data (e.g., user reviews) into rich feature vectors, while a heterogeneous user-product graph models interactions and social ties. Through a tailored message-passing mechanism, text and graph information are fused within the GNN to jointly optimize embeddings. Experiments on public and real-world financial datasets show our model outperforms standalone LLM or GNN in accuracy, recall, and NDCG, with strong interpretability. This work offers new insights for personalized financial recommendations and cross-modal fusion in broader recommendation tasks.
- Open Access
- 1 January 2025
- Published by IGI Global Scientific Publishing in Journal of Organizational and End User Computing
- Vol. 37 (1) , 1-24
Research on Short-Video Platform User Decision-Making via Multimodal Temporal Modeling and Reinforcement Learning
This paper proposes the MT-DQN model, which integrates a Transformer, Temporal Graph Neural Network (TGNN), and Deep Q-Network (DQN) to address the challenges of predicting user behavior and optimizing recommendation strategies in short-video environments. Experiments demonstrated that MT-DQN consistently outperforms traditional concatenated models, such as Concat-Modal, achieving an average F1-score improvement of 10.97% and an average NDCG@5 improvement of 8.3%. Compared to the classic reinforcement learning model Vanilla-DQN, MT-DQN reduces MSE by 34.8% and MAE by 26.5%. Nonetheless, the authors also recognize challenges in deploying MT-DQN in real-world scenarios, such as its computational cost and latency sensitivity during online inference, which will be addressed through future architectural optimization.
- 1 January 2025
- Published by IGI Global Scientific Publishing in Journal of Organizational and End User Computing
- Vol. 37 (1) , 1-24
Open AccessResearch on Short-Video Platform User Decision-Making via Multimodal Temporal Modeling and Reinforcement Learning
This paper proposes the MT-DQN model, which integrates a Transformer, Temporal Graph Neural Network (TGNN), and Deep Q-Network (DQN) to address the challenges of predicting user behavior and optimizing recommendation strategies in short-video environments. Experiments demonstrated that MT-DQN consistently outperforms traditional concatenated models, such as Concat-Modal, achieving an average F1-score improvement of 10.97% and an average NDCG@5 improvement of 8.3%. Compared to the classic reinforcement learning model Vanilla-DQN, MT-DQN reduces MSE by 34.8% and MAE by 26.5%. Nonetheless, the authors also recognize challenges in deploying MT-DQN in real-world scenarios, such as its computational cost and latency sensitivity during online inference, which will be addressed through future architectural optimization.