Probabilistic Coarsening for Knowledge Graph Embeddings
Abstract
:1. Introduction
 Coarsening reduces knowledge graph size whilst preserving the global structure, potentially revealing higherorder features.
 Training schemes that rely on stochastic gradient descent may learn embeddings that fall in local minima. Initializations learned on the coarse graph may be more resistant to this problem.
 Structurally equivalent entities are embedded jointly in coarse graphs, reducing training complexity.
2. Related Work
3. Proposed Strategy
 Probabilistic graph coarsening reduces the base graph to a smaller, coarsened graph and returns an entity mapping between the two graphs.
 Coarse graph embedding applies a predetermined embedding method on the coarse graph to obtain coarse embeddings.
 Reverse mapping and finetuning maps coarse embeddings back down to the base graph to obtain base embeddings. Base embeddings may be finetuned on the base graph.
3.1. Probabilistic Graph Coarsening
3.1.1. Collapsing FirstOrder Neighbours
3.1.2. Collapsing SecondOrder Neighbours
Algorithm 1: Coarse knowledge graph embeddings. 
Input: base graph $\mathcal{K}$; collapsing threshold $\alpha $; random walk count $\eta $ Output: base embeddings $\mathbf{E}$

3.1.3. Neighbour Sampling
 Entities that meet the criteria for collapsing are likely to have smaller neighbourhoods.
 Entities that belong to smaller neighbourhoods have a higher chance of getting sampled as candidates for collapsing.
3.2. Coarse Graph Embedding
3.3. Reverse Mapping and Fine Tuning
4. Evaluation
4.1. Datasets
 MUTAG depicts the properties and interactions of molecules that may or may not be carcinogenic. We remove the labelling predicate isMutagenic from the dataset.
 AIFB reports the work performed at the AIFB research group and labels its members by affiliation. We remove predicates, employs, and affiliation.
 BGS captures geological data from the island of Great Britain and is used to predict the lithogenicity of rocks. As such, we remove the hasLithogenesis predicate.
 AM describes and categorises artefacts in the Amsterdam Museum. We remove the materials predicate as it correlates with artefact labels.
Dataset  MUTAG  AIFB  BGS  AM 

Triples  74,227  29,043  916,199  5,988,321 
Entities  23,644  8285  333,845  1,666,764 
Predicates  23  45  103  133 
Labelled  340  176  146  1000 
Classes  2  4  2  11 
4.2. Procedure
4.3. Results
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Data Availability Statement
Conflicts of Interest
References
 Bordes, A.; Usunier, N.; Chopra, S.; Weston, J. Largescale simple question answering with memory networks. arXiv 2015, arXiv:1506.02075. [Google Scholar]
 Das, R.; Dhuliawala, S.; Zaheer, M.; Vilnis, L.; Durugkar, I.; Krishnamurthy, A.; Smola, A.; McCallum, A. Go for a walk and arrive at the answer: Reasoning over paths in knowledge bases using reinforcement learning. arXiv 2017, arXiv:1711.05851. [Google Scholar]
 Schlichtkrull, M.; Kipf, T.N.; Bloem, P.; Van Den Berg, R.; Titov, I.; Welling, M. Modeling relational data with graph convolutional networks. In European Semantic Web Conference; Springer: Cham, Switzerland, 2018; pp. 593–607. [Google Scholar]
 Bordes, A.; Usunier, N.; GarciaDuran, A.; Weston, J.; Yakhnenko, O. Translating embeddings for modeling multirelational data. Adv. Neural Inf. Process. Syst. 2013, 26, 2787–2795. [Google Scholar]
 Dettmers, T.; Minervini, P.; Stenetorp, P.; Riedel, S. Convolutional 2d knowledge graph embeddings. arXiv 2017, arXiv:1707.01476. [Google Scholar] [CrossRef]
 Bellini, V.; Schiavone, A.; Di Noia, T.; Ragone, A.; Di Sciascio, E. Knowledgeaware autoencoders for explainable recommender systems. In Proceedings of the 3rd Workshop on Deep Learning for Recommender Systems, Vancouver, BC, Canada, 6 October 2018. [Google Scholar]
 Ristoski, P.; Paulheim, H. Rdf2vec: Rdf graph embeddings for data mining. In International Semantic Web Conference; Springer: Cham, Switzerland, 2016; pp. 498–514. [Google Scholar]
 Nickel, M.; Tresp, V.; Kriegel, H.P. A threeway model for collective learning on multirelational data. In Proceedings of the 28th International Conference on Machine Learning, Bellevue, WA, USA, 28 June–2 July 2011. [Google Scholar]
 Chen, H.; Perozzi, B.; Hu, Y.; Skiena, S. Harp: Hierarchical representation learning for networks. In Proceedings of the 32nd AAAI Conference on Artificial Intelligence, New Orleans, LA, USA, 2–7 February 2018; pp. 2127–2134. [Google Scholar]
 Liang, J.; Gurukar, S.; Parthasarathy, S. Mile: A multilevel framework for scalable graph embedding. arXiv 2018, arXiv:1802.09612. [Google Scholar] [CrossRef]
 Archdeacon, D. Topological graph theory. Surv. Congr. Numer. 1996, 115, 18. [Google Scholar]
 Perozzi, B.; AlRfou, R.; Skiena, S. Deepwalk: Online learning of social representations. In Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY, USA, 24–27 August 2014; pp. 701–710. [Google Scholar]
 Tang, J.; Qu, M.; Wang, M.; Zhang, M.; Yan, J.; Mei, Q. Line: Largescale information network embedding. In Proceedings of the 24th International Conference on World Wide Web, Florence, Italy, 18–22 May 2015. [Google Scholar]
 Grover, A.; Leskovec, J. node2vec: Scalable feature learning for networks. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 13–17 August 2016; pp. 855–864. [Google Scholar]
 Mikolov, T.; Sutskever, I.; Chen, K.; Corrado, G.S.; Dean, J. Distributed representations of words and phrases and their compositionality. Adv. Neural Inf. Process. Syst. 2013, 26, 3111–3119. [Google Scholar]
 Duvenaud, D.K.; Maclaurin, D.; Iparraguirre, J.; Bombarell, R.; Hirzel, T.; AspuruGuzik, A.; Adams, R.P. Convolutional networks on graphs for learning molecular fingerprints. Adv. Neural Inf. Process. Syst. 2015, 28, 2224–2232. [Google Scholar]
 Kipf, T.N.; Welling, M. Semisupervised classification with graph convolutional networks. arXiv 2016, arXiv:1609.02907. [Google Scholar]
 Wu, Z.; Pan, S.; Chen, F.; Long, G.; Zhang, C.; Philip, S.Y. A comprehensive survey on graph neural networks. IEEE Trans. Neural Netw. Learn. Syst. 2020, 32, 4–24. [Google Scholar] [CrossRef][Green Version]
 Simonovsky, M.; Komodakis, N. Graphvae: Towards generation of small graphs using variational autoencoders. In International Conference on Artificial Neural Networks; Springer: Cham, Switzerland, 2018; pp. 412–422. [Google Scholar]
 Akyildiz, T.A.; Aljundi, A.A.; Kaya, K. Gosh: Embedding big graphs on small hardware. In Proceedings of the 49th International Conference on Parallel Processing (ICPP), Edmonton, AB, Canada, 17–20 August 2020; pp. 1–11. [Google Scholar]
 Karypis, G.; Kumar, V. Multilevelkway partitioning scheme for irregular graphs. J. Parallel Distrib. Comput. 1998, 48, 96–129. [Google Scholar] [CrossRef][Green Version]
 Wang, Y.; Dong, L.; Jiang, X.; Ma, X.; Li, Y.; Zhang, H. KG2Vec: A node2vecbased vectorization model for knowledge graph. PLoS ONE 2021, 16, e0248552. [Google Scholar] [CrossRef]
 Fionda, V.; Pirró, G. Triple2Vec: Learning Triple Embeddings from Knowledge Graphs. In Proceedings of the AAAI Conference on Artificial Intelligence, New York, NY, USA, 7–12 February 2020. [Google Scholar]
 Portisch, J.; Paulheim, H. Putting rdf2vec in order. In Proceedings of the International Semantic Web Conference (ISWC 2021): Posters and Demo, Virtual Conference, 24–28 October 2021. [Google Scholar]
 Busbridge, D.; Sherburn, D.; Cavallo, P.; Hammerla, N.Y. Relational graph attention networks. arXiv 2019, arXiv:1904.05811. [Google Scholar]
 Yasunaga, M.; Ren, H.; Bosselut, A.; Liang, P.; Leskovec, J. QAGNN: Reasoning with Language Models and Knowledge Graphs for Question Answering. In 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies; Association for Computational Linguistics: Stroudsburg, PA, USA, 2021; pp. 535–546. [Google Scholar]
 Alshahrani, M.; Thafar, M.A.; Essack, M. Application and evaluation of knowledge graph embeddings in biomedical data. PeerJ Comput. Sci. 2021, 7, e341. [Google Scholar] [CrossRef]
 Wang, Z.; Zhang, J.; Feng, J.; Chen, Z. Knowledge graph embedding by translating on hyperplanes. In Proceedings of the AAAI Conference on Artificial Intelligence, Portsmouth, NH, USA, 21–26 June 2014; Volume 28. [Google Scholar]
 Lin, Y.; Liu, Z.; Sun, M.; Liu, Y.; Zhu, X. Learning entity and relation embeddings for knowledge graph completion. In Proceedings of the AAAI Conference on Artificial Intelligence, Austin, TX, USA, 25–30 January 2015; Volume 29. [Google Scholar]
 Ji, G.; He, S.; Xu, L.; Liu, K.; Zhao, J. Knowledge graph embedding via dynamic mapping matrix. In 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers); Association for Computational Linguistics: Stroudsburg, PA, USA, 2015; pp. 687–696. [Google Scholar]
 Xiao, H.; Huang, M.; Hao, Y.; Zhu, X. TransA: An adaptive approach for knowledge graph embedding. arXiv 2015, arXiv:1509.05490. [Google Scholar]
 Nguyen, D.Q.; Sirts, K.; Qu, L.; Johnson, M. STransE: A novel embedding model of entities and relationships in knowledge bases. In 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies; Association for Computational Linguistics: Stroudsburg, PA, USA, 2016; pp. 460–466. [Google Scholar]
 Ebisu, T.; Ichise, R. Toruse: Knowledge graph embedding on a lie group. In Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, LA, USA, 2–7 February 2018; Volume 32. [Google Scholar]
 Sun, Z.; Deng, Z.H.; Nie, J.Y.; Tang, J. RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space. In Proceedings of the International Conference on Learning Representations, New Orleans, LA, USA, 6–9 May 2019. [Google Scholar]
 Yang, B.; Yih, W.t.; He, X.; Gao, J.; Deng, L. Embedding entities and relations for learning and inference in knowledge bases. arXiv 2014, arXiv:1412.6575. [Google Scholar]
 Nickel, M.; Rosasco, L.; Poggio, T. Holographic embeddings of knowledge graphs. In Proceedings of the AAAI Conference on Artificial Intelligence, Phoenix, AZ, USA, 12–17 February 2016; Volume 30. [Google Scholar]
 Balazevic, I.; Allen, C.; Hospedales, T. TuckER: Tensor Factorization for Knowledge Graph Completion. In 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLPIJCNLP); Association for Computational Linguistics: Stroudsburg, PA, USA, 2019; pp. 5185–5194. [Google Scholar]
 Ji, S.; Pan, S.; Cambria, E.; Marttinen, P.; Yu, P.S. A survey on knowledge graphs: Representation, acquisition and applications. arXiv 2020, arXiv:2002.00388. [Google Scholar] [CrossRef]
 Pietrasik, M.; Reformat, M. A Simple Method for Inducing Class Taxonomies in Knowledge Graphs. In European Semantic Web Conference; Springer: Cham, Switzerland, 2020; pp. 53–68. [Google Scholar]
 Hendrickson, B.; Leland, R.W. A MultiLevel Algorithm For Partitioning Graphs. SC 1995, 95, 1–14. [Google Scholar]
 Karypis, G.; Kumar, V. A fast and high quality multilevel scheme for partitioning irregular graphs. SIAM J. Sci. Comput. 1998, 20, 359–392. [Google Scholar] [CrossRef]
 Han, X.; Cao, S.; Xin, L.; Lin, Y.; Liu, Z.; Sun, M.; Li, J. OpenKE: An Open Toolkit for Knowledge Embedding. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, Brussels, Belgium, 31 October–4 November 2018. [Google Scholar]
 Portisch, J.; Hladik, M.; Paulheim, H. RDF2Vec Light—A Lightweight Approach for Knowledge Graph Embeddings. In Proceedings of the International Semantic Web Conference, Posters and Demos, Virtual Conference, 1–6 November 2020. [Google Scholar]
 Portisch, J.; Paulheim, H. Walk this way! entity walks and property walks for rdf2vec. In The Semantic Web: ESWC 2022 Satellite Events: Hersonissos, Crete, Greece, 29 May–2 June 2022, Proceedings; Springer: Cham, Switerland, 2022; pp. 133–137. [Google Scholar]
 Cochez, M.; Ristoski, P.; Ponzetto, S.P.; Paulheim, H. Biased graph walks for RDF graph embeddings. In Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, Amantea, Italy, 19–22 June 2017; pp. 1–12. [Google Scholar]
 Portisch, J.; Heist, N.; Paulheim, H. Knowledge graph embedding for data mining vs. knowledge graph embedding for link prediction–two sides of the same coin? Semant. Web 2022, 13, 399–422. [Google Scholar] [CrossRef]
 Bhatt, S.; Padhee, S.; Sheth, A.; Chen, K.; Shalin, V.; Doran, D.; Minnery, B. Knowledge graph enhanced community detection and characterization. In Proceedings of the twelfth ACM International Conference on Web Search and Data Mining, Melbourne, VIC, Australia, 11–15 February 2019; pp. 51–59. [Google Scholar]
 Shi, X.; Qian, Y.; Lu, H. Community Detection in Knowledge Graph Network with Matrix Factorization Learning. In Web and Big Data: APWebWAIM 2019 International Workshops, KGMA and DSEA, Chengdu, China, August 1–3, 2019, Revised Selected Papers 3; Springer: Cham, Switzerland, 2019; pp. 37–51. [Google Scholar]
 Paul, S.; Chen, Y. Consistent community detection in multirelational data through restricted multilayer stochastic blockmodel. Electron. J. Stat. 2016, 10, 3807–3870. [Google Scholar] [CrossRef]
 De Bacco, C.; Power, E.A.; Larremore, D.B.; Moore, C. Community detection, link prediction, and layer interdependence in multilayer networks. Phys. Rev. E 2017, 95, 042317. [Google Scholar] [CrossRef] [PubMed][Green Version]
Method  MUTAG  AIFB  BGS  AM 

RDF2Vec  $0.7500\pm 0.0392$  $0.9111\pm 0.0117$  $0.7828\pm 0.0327$  $0.8758\pm 0.0143$ 
C(RDF2Vec)  $\underline{\mathbf{0}.\mathbf{7956}\pm \mathbf{0}.\mathbf{0340}}$  $\mathbf{0}.\mathbf{9167}\pm \mathbf{0}.\mathbf{0000}$  $\underline{\mathbf{0}.\mathbf{8828}\pm \mathbf{0}.\mathbf{0178}}$  $\mathbf{0}.\mathbf{8778}\pm \mathbf{0}.\mathbf{0211}$ 
Change  6.1% *  0.6%  12.8% *  0.2% 
RGCN  $\mathbf{0}.\mathbf{7397}\pm \mathbf{0}.\mathbf{0286}$  $0.9528\pm 0.0264$  $0.8345\pm 0.0424$  $\underline{\mathbf{0}.\mathbf{8833}\pm \mathbf{0}.\mathbf{0197}}$ 
C(RGCN)  $0.7294\pm 0.0242$  $\underline{\mathbf{0}.\mathbf{9694}\pm \mathbf{0}.\mathbf{0088}}$  $\mathbf{0}.\mathbf{8690}\pm \mathbf{0}.\mathbf{0317}$  $0.8828\pm 0.0138$ 
Change  −1.4%  1.7% *  4.1% *  −0.1% 
TransE  $0.7397\pm 0.0422$  $0.8722\pm 0.0397$  $0.6793\pm 0.0371$  $0.4207\pm 0.0143$ 
C(TransE)  $\mathbf{0}.\mathbf{7412}\pm \mathbf{0}.\mathbf{0368}$  $\mathbf{0}.\mathbf{9056}\pm \mathbf{0}.\mathbf{0299}$  $\mathbf{0}.\mathbf{7759}\pm \mathbf{0}.\mathbf{0335}$  $\mathbf{0}.\mathbf{4955}\pm \mathbf{0}.\mathbf{0179}$ 
Change  0.3%  3.8% *  14.2% *  17.7% * 
Dataset  MUTAG  AIFB  BGS  AM 

Triples  52179  20134  501722  4080981 
Change  −29.7%  −30.7%  −45.2%  −31.8% 
Entities  16115  2801  78335  944759 
Change  −31.8%  −66.2%  −76.5%  −43.3% 
Predicates  23  43  97  129 
Change  0%  −4.4%  −5.8%  −3.0% 
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. 
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Pietrasik, M.; Reformat, M.Z. Probabilistic Coarsening for Knowledge Graph Embeddings. Axioms 2023, 12, 275. https://doi.org/10.3390/axioms12030275
Pietrasik M, Reformat MZ. Probabilistic Coarsening for Knowledge Graph Embeddings. Axioms. 2023; 12(3):275. https://doi.org/10.3390/axioms12030275
Chicago/Turabian StylePietrasik, Marcin, and Marek Z. Reformat. 2023. "Probabilistic Coarsening for Knowledge Graph Embeddings" Axioms 12, no. 3: 275. https://doi.org/10.3390/axioms12030275