何铁科，南京大学准聘副教授（特聘研究员）、博导。

艾塔实验室主页：https://eta.nju.edu.cn/

邮箱：hetieke {AT} nju DOT edu DOT cn, hetieke {AT} gmail DOT com

招募优秀本科生加入艾塔实验室，参与大模型与智能系统研究，开启高水平科研探索之旅。

研究方向包括：大模型、Agentic AI、图机器学习、大数据与工业智能应用 等。

团队坚持理论深度与工程实践并重，提供稳定科研经费、充足算力资源及国际合作与学术交流机会。

教学：

本科通识课《AI辅助工作》

本科二年级《数据库系统概论》

本科三年级《大数据分析》

研究生《统计模型：理论与实践》

研究生《大数据技术与应用》

代表论著：

Bridging Academia and Industry: A Comprehensive Benchmark for Attributed Graph Clustering. NeurIPS'26

Uncovering Semantic Hierarchies in Text-Attributed Graphs via Variational EM-based LLM–GNN Synergy. NeurIPS'26

HELO-APR: Enhancing Low-Resource Program Repair through Cross-Lingual Knowledge Transfer. ISSTA'26

GS-Quant: Granular Semantic and Generative Structural Quantization for Knowledge Graph Completion. ACL'26

A Survey of Reinforcement Learning for Large Language Models under Data Scarcity: Challenges and Solutions. ACL'26

CoDA: Restoring Contextual Dominance via Copy-Encouraged Attention Intervention for Mitigating RAG Hallucinations. ACL'26

Towards Anomaly-Aware Pre-Training and Fine-Tuning for Graph Anomaly Detection. ICLR'26

Tabular Foundation Models are Strong Graph Anomaly Detectors. WWW'26

Enhancing Multi-Modal Entity Alignment via Multi-Grained Decision Fusion. WWW'26

Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive Approach. WWW'26

Learning Hierarchical Knowledge in Text-Rich Networks with Taxonomy-Informed Representation Learning. KDD'26

On Modality Weighting and Specificity for Multi-Modal Entity Alignment. AAAI'26

EchoEdit: Consistent Multi-Hop Question Answering via Ripple Control in Knowledge Editing. AAAI'26

Norm-Augmented Graph AutoEncoders for Link Prediction. ICASSP'25

Dynamic Graph Recommendation via Sparse Augmentation and Singular Adaptation. ICASSP'25

Multi-Scale Heterogeneous Text-Attributed Graph Datasets From Diverse Domains. WWW'25

Learning Efficient and Accurate MLPs on Multiplex Graphs via Ensemble Distillation. DASFAA'25

Scalable and Adaptive Spectral Embedding for Attributed Graph Clustering. CIKM'24

Bootstrap Latents of Nodes and Neighbors for Graph Self-Supervised Learning. ECML'24

Interaction-level Membership Inference Attack Against Federated Recommender Systems. WWW'23

Federated Unlearning for On-Device Recommendation. WSDM'22

CIRCLE: Continual Repair across Programming Languages. ISSTA'22

Unified Question Generation with Continual Lifelong Learning. WWW'22

Improving Neural Question Generation using Deep Linguistic Representation. WWW'21

Crowdsourced Requirements Generation for Automatic Testing via Knowledge Graph. ISSTA'20

━━━━✧❂✧━━━━✧❂✧━━━━✧❂✧━━━━✧❂✧━━━━✧❂✧━━━━✧❂✧━━━━✧❂✧━━━━

[J] Teaching MLPs to Master Hetero-Graph-Structured Knowledge for Efficient and Accurate Inference. TKDE (2025)

[J] 大模型在代码优化任务中的能力探究及改进方法. 《软件学报》(2024)

[J] Reliable Node Similarity Matrix Guided Contrastive Graph Clustering. TKDE (2024)

[J] Negative-Free Self-Supervised Gaussian Embedding of Graphs. Neural Networks (2024)

[J] APPT: Boosting Automated Patch Correctness Prediction via Fine-tuning Pre-trained Models. TSE (2024)

[J] A Systemic Framework for Crowdsourced Test Report Quality Assessment. EMSE (2020)

[J] PTM: A Topic Model for the Inferring of the Penalty. JCST (2018)