請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/101287完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 林澤 | zh_TW |
| dc.contributor.advisor | Che Lin | en |
| dc.contributor.author | 林少恩 | zh_TW |
| dc.contributor.author | Shao-En Lin | en |
| dc.date.accessioned | 2026-01-13T16:13:22Z | - |
| dc.date.available | 2026-01-14 | - |
| dc.date.copyright | 2026-01-13 | - |
| dc.date.issued | 2025 | - |
| dc.date.submitted | 2025-12-16 | - |
| dc.identifier.citation | T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama. Optuna: A next-generation hyperparameter optimization framework. In The 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pages 2623–2631, 2019.
L. Butler, A. Parada-Mayorga, and A. Ribeiro. Convolutional learning on multigraphs. IEEE Transactions on Signal Processing, 71:933–946, Jan. 2023. ISSN 1053-587X. doi:10.1109/TSP.2023.3259144. I. Cantador, P. Brusilovsky, and T. Kuflik. Second workshop on information heterogeneity and fusion in recommender systems (HetRec2011). In Proceedings of the Fifth ACM Conference on Recommender Systems, pages 387–388, New York, NY, USA, 2011. Association for Computing Machinery. ISBN 9781450306836. doi:10.1145/2043932.2044016. Y. Cen, X. Zou, J. Zhang, H. Yang, J. Zhou, and J. Tang. Representation learning for attributed multiplex heterogeneous network. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 1358–1368, ACM, 2019. D. Chen, Y. Lin, W. Li, P. Li, J. Zhou, and X. Sun. Measuring and relieving the over-smoothing problem for graph neural networks from the topological view. Proceedings of the AAAI Conference on Artificial Intelligence, 34(04):3438–3445, Apr. 2020. doi:10.1609/aaai.v34i04.5747. K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan. A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation, 6(2):182–197, 2002. doi:10.1109/4235.996017. M. Defferrard, X. Bresson, and P. Vandergheynst. Convolutional neural networks on graphs with fast localized spectral filtering. In Advances in Neural Information Processing Systems, 2016. M. Fey, W. Hu, K. Huang, J. E. Lenssen, R. Ranjan, J. Robinson, R. Ying, J. You, and J. Leskovec. Position: Relational deep learning – graph representation learning on relational databases. In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceedings of Machine Learning Research, pages 13592–13607, PMLR, Jul. 2024. X. Fu, J. Zhang, Z. Meng, and I. King. MAGNN: Metapath aggregated graph neural network for heterogeneous graph embedding. In Proceedings of The Web Conference 2020, pages 2331–2341, New York, NY, USA, 2020. Association for Computing Machinery. W. L. Hamilton, R. Ying, and J. Leskovec. Inductive representation learning on large graphs. In Advances in Neural Information Processing Systems, 2017. M. He, Z. Wei, S. Feng, Z. Huang, W. Li, Y. Sun, and D. Yu. Spectral heterogeneous graph convolutions via positive noncommutative polynomials. In Proceedings of the ACM Web Conference 2024, pages 685–696, New York, NY, USA, 2024. Association for Computing Machinery. X. He, K. Deng, X. Wang, Y. Li, Y. Zhang, and M. Wang. LightGCN: Simplifying and powering graph convolution network for recommendation. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 639–648, 2020. W. Hu, M. Fey, M. Zitnik, Y. Dong, H. Ren, B. Liu, M. Catasta, and J. Leskovec. Open Graph Benchmark: Datasets for machine learning on graphs. In Advances in Neural Information Processing Systems, 2020. Z. Hu, Y. Dong, K. Wang, and Y. Sun. Heterogeneous graph transformer. In Proceedings of The Web Conference 2020, pages 2704–2710, 2020. S. Katoch, S. S. Chauhan, and V. Kumar. A review on genetic algorithm: past, present, and future. Multimedia Tools and Applications, 80(5):8091–8126, 2021. T. N. Kipf and M. Welling. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations, 2017. C. Li, Z. Guo, Q. He, H. Xu, and K. He. Long-range meta-path search on large-scale heterogeneous graphs, 2024. Q. Lv, M. Ding, Q. Liu, Y. Chen, W. Feng, S. He, C. Zhou, J. Jiang, Y. Dong, and J. Tang. Are we really making much progress? Revisiting, benchmarking and refining heterogeneous graph neural networks. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 1150–1160, 2021. Q. Mao, Z. Liu, C. Liu, and J. Sun. HINormer: Representation learning on heterogeneous information networks with graph transformer. In Proceedings of the ACM Web Conference 2023, pages 599–610, 2023. Y. Ozaki, Y. Tanigaki, S. Watanabe, and M. Onishi. Multiobjective tree-structured Parzen estimator for computationally expensive optimization problems. In Proceedings of the 2020 Genetic and Evolutionary Computation Conference, pages 533–541, 2020. M. Schlichtkrull, T. N. Kipf, P. Bloem, R. van den Berg, I. Titov, and M. Welling. Modeling relational data with graph convolutional networks. In The Semantic Web, pages 593–607, Springer, 2018. Y. Sun, J. Han, X. Yan, P. S. Yu, and T. Wu. PathSim: Meta path-based top-k similarity search in heterogeneous information networks. Proceedings of the VLDB Endowment, 4(11):992–1003, 2011. P. Veličković, G. Cucurull, A. Casanova, A. Romero, P. Liò, and Y. Bengio. Graph Attention Networks. International Conference on Learning Representations, 2018. J. Wang, Y. Guo, L. Yang, and Y. Wang. Enabling homogeneous GNNs to handle heterogeneous graphs via relation embedding. IEEE Transactions on Big Data, 9:1697–1710, 2023. M. Wang et al. 4DBInfer: A 4D benchmarking toolbox for graph-centric predictive modeling on relational databases. In Advances in Neural Information Processing Systems, volume 37, pages 27236–27273, 2024. X. Wang, H. Ji, C. Shi, B. Wang, Y. Ye, P. Cui, and P. S. Yu. Heterogeneous graph attention network. In The World Wide Web Conference, pages 2022–2032, 2019. F. Wu, A. Souza, T. Zhang, C. Fifty, T. Yu, and K. Weinberger. Simplifying graph convolutional networks. In Proceedings of the 36th International Conference on Machine Learning, pages 6861–6871, 2019. B. Yang, W. T. Yih, X. He, J. Gao, and L. Deng. Embedding entities and relations for learning and inference in knowledge bases. International Conference on Learning Representations, 2014. C. Yang, Y. Xiao, Y. Zhang, Y. Sun, and J. Han. Heterogeneous network representation learning: A unified framework with survey and benchmark. IEEE Transactions on Knowledge and Data Engineering, 2020. X. Yang, M. Yan, S. Pan, X. Ye, and D. Fan. Simple and efficient heterogeneous graph neural network. Proceedings of the AAAI Conference on Artificial Intelligence, 37(9):10816–10824, 2023. P. Yu, C. Fu, Y. Yu, C. Huang, Z. Zhao, and J. Dong. Multiplex heterogeneous graph convolutional network. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 2377–2387, 2022. S. Yun, M. Jeong, R. Kim, J. Kang, and H. J. Kim. Graph transformer networks. In Advances in Neural Information Processing Systems, pages 11960–11970, 2019. Z. Zhou, J. Shi, R. Yang, Y. Zou, and Q. Li. SlotGAT: Slot-based message passing for heterogeneous graphs. In Proceedings of the 40th International Conference on Machine Learning, 2023. | - |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/101287 | - |
| dc.description.abstract | 異質資訊網路（Heterogeneous Information Networks, HINs）提供了一個具有表現力的框架，能夠整合不同類型的實體與關係，通常定義於一個固定的綱要（schema）之下。然而，現有研究多倚賴預設的綱要結構，忽略了資料在不同建模選擇下呈現的差異，進而對下游任務造成潛在影響。
本研究提出「實體與屬性二象性（entity–attribute duality）」的概念：屬性可以被原子化為具關聯的實體，而實體亦可作為其他實體的屬性。此原則啟發我們設計出 atomic HIN 的範式，旨在系統性地規範結構建模選擇，並提升異質圖模型的表達能力。 在此基礎上，我們進一步提出一個任務導向的綱要最佳化（schema refinement）框架，能將既有資料集視為某一特定設計選擇下的產物，並系統性地探索更適合下游任務的綱要結構。實驗結果顯示，在結合 schema refinement 後，僅使用一個簡化版的關係圖卷積網（simplified Relational GCN, sRGCN），即可於八個涵蓋節點層級與邊層級任務的資料集上達到當前最佳（state-of-the-art）表現；進一步使用先進的異質圖神經網路（HGNNs）則可進一步提升效果。這表明，異質圖的結構建模選擇對於學習表現具有關鍵性影響。 最後，我們公開 atomic HINs 的最佳化結構與完整框架，為建立更具原則性的模型評分奠定基礎，並為未來在綱要感知式學習（schema-aware learning）、自動結構探索（automated structure discovery）以及下一代 HGNNs 的研究提供新方向。 | zh_TW |
| dc.description.abstract | Heterogeneous Information Networks (HINs) provide a powerful framework for modeling multi-typed entities and relations, typically defined under a fixed schema. Yet, most research assumes this structure is given, overlooking the fact that alternative designs can emphasize different aspects of the data and substantially influence downstream performance.
As a theoretical foundation for such designs, we introduce the principle of entity-attribute duality: attributes can be atomized as entities with their associated relations, while entities can, in turn, serve as attributes of others. This principle motivates atomic HIN, a canonical representation that makes all modeling choices explicit and achieves maximal expressiveness. Building on this foundation, we propose a systematic framework for task-specific schema refinement. Within this framework, we demonstrate that widely used benchmarks correspond to heuristic refinements of the atomic HIN—often far from optimal. Across eight datasets, refinement alone enables a simplified Relational GCN (sRGCN) to reach state-of-the-art performance on node- and link-level tasks, with further gains from advanced HGNNs. These results highlight schema design as a key dimension in heterogeneous graph modeling. By releasing the atomic HINs, searched schemas, and refinement framework, we enable principled benchmarking and open the way for future work on schema-aware learning, automated structure discovery, and next-generation HGNNs. | en |
| dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2026-01-13T16:13:22Z No. of bitstreams: 0 | en |
| dc.description.provenance | Made available in DSpace on 2026-01-13T16:13:22Z (GMT). No. of bitstreams: 0 | en |
| dc.description.tableofcontents | Acknowledgements i
摘要 iii Abstract v Contents vii List of Figures xi List of Tables xiii Chapter 1 Introduction 1 Section 1.1 Motivation 1 Section 1.2 Related Work 4 Section 1.2.1 Heterogeneous Graph Neural Networks 4 Section 1.2.2 Schema Augmentation Techniques 5 Section 1.2.3 Schema Design in Common Benchmarks 6 Chapter 2 Methodology 9 Section 2.1 Preliminaries 10 Section 2.1.1 Attributed Heterogeneous Information Networks 10 Section 2.1.2 Spectral Heterogeneous Graph Convolution (SHGC) 11 Section 2.2 From Attribute Atomization to Atomic HINs 12 Section 2.3 Schema Refinement via Node- and Edge-Type Selection 14 Section 2.4 Pre-propagation Feature Initialization 16 Section 2.5 Systematic Search for Schema Refinement 19 Section 2.6 HGNNs for Atomic HINs 20 Chapter 3 Simplified Relational GCN 27 Section 3.1 Formulation 28 Section 3.2 Layer-wise Update Rule 28 Section 3.3 Precomputation Setting 30 Section 3.4 Downstream Decoders 30 Section 3.5 Node Classification 31 Section 3.6 Link Prediction 31 Section 3.7 Training Enhancements 32 Section 3.8 L2-normalization 32 Section 3.9 Edge Dropout and Input Dropout 33 Section 3.10 Residual Connection 33 Section 3.11 Label Propagation 34 Section 3.12 Loss Function 35 Chapter 4 Experiments 37 Section 4.1 Datasets 37 Section 4.1.1 Human-Designed Schemas as Specific Refinements 38 Section 4.1.2 IMDB 39 Section 4.1.3 Freebase 40 Section 4.1.4 DBLP 41 Section 4.1.5 ACM 41 Section 4.1.6 OGBN-MAG 42 Section 4.1.7 Amazon 43 Section 4.1.8 LastFM 43 Section 4.1.9 PubMed 44 Section 4.2 Evaluation Protocol 45 Section 4.2.1 Evaluation Metrics 45 Section 4.2.1.1 Confusion Matrix 46 Section 4.2.1.2 Accuracy 46 Section 4.2.1.3 F1-Scores 47 Section 4.2.1.4 AUC-ROC 48 Section 4.2.1.5 MRR 49 Section 4.2.2 Dataset Split 49 Section 4.2.3 Negative Sampling 50 Section 4.3 Baselines 51 Section 4.4 Schema Refinement and Hyperparameter Setting 53 Section 4.4.1 Schema Tuning 53 Section 4.4.2 HGNN Hyperparameters and Tuning 53 Section 4.5 Results 54 Section 4.5.1 Results on Node Classification Datasets 54 Section 4.5.2 Results on Link Prediction Datasets 56 Chapter 5 Discussion 59 Section 5.1 Search Results on sRGCN 59 Section 5.2 Effect of Attribute Atomization and Schema Refinement 63 Section 5.3 Transferability and Generalization of Refined Schemas 64 Section 5.4 Search Results on More HGNNs 66 Section 5.5 Efficiency of Schema Refinement 69 Section 5.6 Discussion on Search Algorithms and Initialization 69 Section 5.7 Complexity Analysis 71 Chapter 6 Conclusion 75 References 79 Appendix A — Details of Pre-propagation Feature Initialization 87 Section A.1 Generalized Union Aggregation 87 Section A.2 Implementation 88 | - |
| dc.language.iso | en | - |
| dc.subject | 異質資訊網路 | - |
| dc.subject | 異質圖神經網路 | - |
| dc.subject | 圖神經網路 | - |
| dc.subject | 圖表徵學習 | - |
| dc.subject | 深度學習 | - |
| dc.subject | Heterogeneous Information Networks | - |
| dc.subject | Heterogeneous Graph Neural Networks | - |
| dc.subject | Graph Neural Networks | - |
| dc.subject | Graph Representation Learning | - |
| dc.subject | Deep Learning | - |
| dc.title | 基於實體-屬性二象性的異質圖建模 | zh_TW |
| dc.title | Entity-Attribute Duality for Heterogeneous Graph Modeling | en |
| dc.type | Thesis | - |
| dc.date.schoolyear | 114-1 | - |
| dc.description.degree | 碩士 | - |
| dc.contributor.oralexamcommittee | 王志宇;王釧茹;蔡銘峰 | zh_TW |
| dc.contributor.oralexamcommittee | Chih-Yu Wang;Chuan-Ju Wang;Ming-Feng Tsai | en |
| dc.subject.keyword | 異質資訊網路,異質圖神經網路圖神經網路圖表徵學習深度學習 | zh_TW |
| dc.subject.keyword | Heterogeneous Information Networks,Heterogeneous Graph Neural NetworksGraph Neural NetworksGraph Representation LearningDeep Learning | en |
| dc.relation.page | 89 | - |
| dc.identifier.doi | 10.6342/NTU202504725 | - |
| dc.rights.note | 同意授權(限校園內公開) | - |
| dc.date.accepted | 2025-12-17 | - |
| dc.contributor.author-college | 電機資訊學院 | - |
| dc.contributor.author-dept | 電信工程學研究所 | - |
| dc.date.embargo-lift | 2027-11-26 | - |
| 顯示於系所單位： | 電信工程學研究所 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-114-1.pdf 未授權公開取用 | 2.26 MB | Adobe PDF | 檢視/開啟 |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。