Can Persistent Homology provide an efficient alternative for Evaluation of Knowledge Graph Completion Methods?
Not on the map. Why?
Abstract (arXiv)
In this paper we present a novel method, textitKnowledge Persistence (KP), for faster evaluation of Knowledge Graph (KG) completion approaches. Current ranking-based evaluation is quadratic in the size of the KG, leading to long evaluation times and consequently a high carbon footprint. KP addresses this by representing the topology of the KG completion methods through the lens of topological data analysis, concretely using persistent homology. The characteristics of persistent homology allow KP to evaluate the quality of the KG completion looking only at a fraction of the data. Experimental results on standard datasets show that the proposed metric is highly correlated with ranking metrics (Hits@N, MR, MRR). Performance evaluation shows that KP is computationally efficient: In some cases, the evaluation time (validation+test) of a KG completion method has been reduced from 18 hours (using Hits@10) to 27 seconds (using KP), and on average (across methods & data) reduces the evaluation time (validation+test) by ≈ 99.96%.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 29 of its 68 references that are in the map. First posted on arXiv 30 Jan 2023.
Builds on
29 of 68 references.
References · 29
- 2014Embedding Entities and Relations for Learning and Inference in Knowledge Basescited by 929
- 2016Complex Embeddings for Simple Link Predictioncited by 897
- 2019RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Spacecited by 893
- 2020A Survey on Knowledge Graphs: Representation, Acquisition and Applicationscited by 643
- 2019KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representationcited by 356
- 2019TuckER: Tensor Factorization for Knowledge Graph Completioncited by 290
- 2021Sustainable AI: Environmental Implications, Challenges and Opportunitiescited by 242
- 2022The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrinkcited by 183
- 2017A Novel Embedding Model for Knowledge Base Completion Based on Convolutional Neural Networkcited by 143
- 2019Topological Autoencoderscited by 124
- 2020CoDEx: A Comprehensive Knowledge Graph Completion Benchmarkcited by 102
- 2017Sliced Wasserstein Kernel for Persistence Diagramscited by 77
17 more
- 2018Neural Persistence: A Complexity Measure for Deep Neural Networks Using Algebraic Topologycited by 73
- 2019A Re-evaluation of Knowledge Graph Completion Methodscited by 73
- 2020Bringing Light Into the Dark: A Large-scale Evaluation of Knowledge Graph Embedding Models Under a Unified Frameworkcited by 70
- 2017Knowledge Base Completion: Baselines Strike Backcited by 39
- 2021Fast Approximation of the Sliced-Wasserstein Distance Using Concentration of Random Projectionscited by 33
- 2019A Survey on Knowledge Graph Embeddings with Literals: Which model links better Literal-ly?cited by 31
- 2019Probability Calibration for Knowledge Graph Embedding Modelscited by 20
- 2022On the effectiveness of persistent homologycited by 20
- 2016Topological Data Analysiscited by 16
- 2021Efficient Non-Sampling Knowledge Graph Embeddingcited by 14
- 2020Knowledge Base Completion: Baseline strikes back (Again)cited by 12
- 2021Highly Efficient Knowledge Graph Embedding Learning with Orthogonal Procrustes Analysiscited by 10
- 2019Pykg2vec: A Python Library for Knowledge Graph Embeddingcited by 9
- 2021HopfE: Knowledge Graph Representation Learning using Inverse Hopf Fibrationscited by 9
- 2022Swift and Sure: Hardness-aware Contrastive Learning for Low-dimensional Knowledge Graph Embeddingscited by 8
- 2021Ranking vs. Classifying: Measuring Knowledge Base Completion Qualitycited by 4
- 2021Topology Applied to Machine Learning: From Global to Localcited by 4
Built on by
Citing papers per year: 2024: 1.
1 paper.