Ms. Surong Yan | Graph-based recommendation systems | Best Researcher Award

Professor, Zhejiang University of Finance and Economics, China📖

Scopus Profile

Education and Experience🎓

Professional Development🌱

Dr. Yan has been a dedicated member of the faculty at Zhejiang University of Finance and Economics since 2005, where she progressed from Lecturer to Associate Professor and ultimately to Professor. In addition to her teaching and research contributions, she spent a year as a visiting scholar at the University of California, Irvine, in 2015-2016, where she furthered her work in data mining and human activity recognition. Her contributions as an academic include serving as a reviewer for prestigious journals such as IEEE Transactions on Knowledge and Data Engineering (TKDE), Knowledge-Based Systems (KBS), Expert Systems with Applications (ESWA), and as a program committee (PC) member for international conferences.

Research Focus🔬

Dr. Surong Yan’s research focuses on data mining, knowledge discovery, artificial intelligence, and recommendation systems. Her work addresses real-world challenges in online activity recognition, human interaction, and recommendation systems in the context of IoT and social computing. Her innovative methodologies leverage edge computing, graph neural networks, and reinforcement learning to enhance intelligent services and human-computer interaction.

Author Metrics 

Publications Top Notes 📄

1. Feature Interactive Graph Neural Network for KG-based Recommendation
   • Authors: Yan, S., Li, C., Wang, H., Lin, B., Yuan, Y.
   • Published in: Expert Systems with Applications
   • Year: 2024
   • Volume: 237
   • Article ID: 121411
   • Abstract: This study proposes a novel feature interactive graph neural network (GIN) model for knowledge graph (KG)-based recommendation systems. It explores feature interactions between different types of data (e.g., user, item, and context) in a knowledge graph to enhance recommendation accuracy. The model uses graph neural networks to process and learn from the structured knowledge embedded in the graph, which helps to improve the personalized recommendation systems.
   • Citations: 3
2. Cross-view Temporal Graph Contrastive Learning for Session-based Recommendation
   • Authors: Wang, H., Yan, S., Wu, C., Han, L., Zhou, L.
   • Published in: Knowledge-Based Systems
   • Year: 2023
   • Volume: 264
   • Article ID: 110304
   • Abstract: This paper introduces a cross-view temporal graph contrastive learning method for session-based recommendation systems. The authors address the challenge of recommending items in a session-based environment, where the user preferences change over time. The proposed model utilizes temporal graph-based contrastive learning to capture these changes and enhance the accuracy of session-based recommendation engines.
   • Citations: 10
3. LkeRec: Toward Lightweight End-To-End Joint Representation Learning for Building Accurate and Effective Recommendation
   • Authors: Yan, S., Lin, K.-J., Zheng, X., Wang, H.
   • Published in: ACM Transactions on Information Systems
   • Year: 2022
   • Volume: 40(3)
   • Article ID: 54
   • Abstract: LkeRec is an end-to-end lightweight joint representation learning approach aimed at improving the accuracy and efficiency of recommendation systems. It combines feature learning and recommendation generation into a single framework to simplify the process and reduce computation costs, while ensuring that the quality of recommendations remains high.
   • Citations: 5
4. A Hybrid Model with Novel Feature Selection Method and Enhanced Voting Method for Credit Scoring
   • Authors: Yao, J., Wang, Z., Wang, L., Jiang, H., Yan, S.
   • Published in: Journal of Intelligent and Fuzzy Systems
   • Year: 2022
   • Volume: 42(3)
   • Pages: 2565–2579
   • Abstract: The paper presents a hybrid model for credit scoring, which integrates a novel feature selection method with an enhanced voting mechanism to improve prediction accuracy. The hybrid model combines different machine learning techniques to identify the most relevant features for credit scoring and make better predictions, particularly in the context of financial applications.
   • Citations: 4
5. Attention-Aware Metapath-Based Network Embedding for HIN-Based Recommendation
   • Authors: Yan, S., Wang, H., Li, Y., Zheng, Y., Han, L.
   • Published in: Expert Systems with Applications
   • Year: 2021
   • Volume: 174
   • Article ID: 114601
   • Abstract: This paper proposes an attention-aware metapath-based network embedding method for heterogeneous information networks (HINs). The proposed method focuses on the importance of different metapaths and incorporates attention mechanisms to improve the recommendation quality in HIN-based environments. The approach is particularly useful for scenarios where the relationships between entities are complex and diverse.
   • Citations: 30

Conclusion