mlsanityData to 14 Oct 2026

Predictive Multiplicity of Knowledge Graph Embeddings in Link Prediction

arXiv 2408.08226 ↗First posted 15 Aug 2024cs.AIcited by 9

Not on the map. Why?

Abstract (arXiv)

Knowledge graph embedding (KGE) models are often used to predict missing links for knowledge graphs (KGs). However, multiple KG embeddings can perform almost equally well for link prediction yet give conflicting predictions for unseen queries. This phenomenon is termed textitpredictive multiplicity in the literature. It poses substantial risks for KGE-based applications in high-stake domains but has been overlooked in KGE research. We define predictive multiplicity in link prediction, introduce evaluation metrics and measure predictive multiplicity for representative KGE methods on commonly used benchmark datasets. Our empirical study reveals significant predictive multiplicity in link prediction, with 8% to 39% testing queries exhibiting conflicting predictions. We address this issue by leveraging voting methods from social choice theory, significantly mitigating conflicts by 66% to 78% in our experiments.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 24 of its 59 references that are in the map. First posted on arXiv 15 Aug 2024.

Builds on

24 of 59 references.

References · 24

  1. 2014Embedding Entities and Relations for Learning and Inference in Knowledge Basescited by 932
  2. 2016Complex Embeddings for Simple Link Predictioncited by 902
  3. 2019RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Spacecited by 898
  4. 2017Convolutional 2D Knowledge Graph Embeddingscited by 730
  5. 2020A Survey on Knowledge Graphs: Representation, Acquisition and Applicationscited by 643
  6. 2020Knowledge Graphscited by 621
  7. 2019Predictive Multiplicity in Classificationcited by 166
  8. 2020BoxE: A Box Embedding Model for Knowledge Base Completioncited by 105
  9. 2020Complex Query Answering with Neural Link Predictorscited by 99
  10. 2018One-Shot Relational Learning for Knowledge Graphscited by 78
  11. 2020On Counterfactual Explanations under Predictive Multiplicitycited by 74
  12. 2019Probabilistic Logic Neural Networks for Reasoningcited by 69
12 more
  1. 2020Interpreting Neural Networks as Gradual Argumentation Frameworks (Including Proof Appendix)cited by 39
  2. 2019Probability Calibration for Knowledge Graph Embedding Modelscited by 21
  3. 2017On Multi-Relational Link Prediction with Bilinear Modelscited by 17
  4. 2021Adversarial Attacks on Knowledge Graph Embeddings via Instance Attribution Methodscited by 16
  5. 2022A Unified Framework for Rank-based Evaluation Metrics for Link Prediction in Knowledge Graphscited by 15
  6. 2023Recourse under Model Multiplicity via Argumentative Ensembling (Technical Report)cited by 15
  7. 2021Poisoning Knowledge Graph Embeddings via Relation Inference Patternscited by 12
  8. 2023Shrinking Embeddings for Hyper-Relational Knowledge Graphscited by 12
  9. 2023Temporal Fact Reasoning over Hyper-Relational Knowledge Graphscited by 10
  10. 2021Multiple Run Ensemble Learning with Low-Dimensional Knowledge Graph Embeddingscited by 8
  11. 2024Generating SROI^- Ontologies via Knowledge Graph Query Embedding Learningcited by 5
  12. 2023Robust Knowledge Extraction from Large Language Models using Social Choice Theorycited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 3; 2026: 5.

9 papers.

  1. 2025Certainty in Uncertainty: Reasoning over Uncertain Knowledge Graphs with Statistical Guaranteescited by 9
  2. 2025Predicate-Conditional Conformalized Answer Sets for Knowledge Graph Embeddingscited by 4
  3. 2026AWM: Answerable Working Memory for Long-Document VQA Agentscited by 2
  4. 2024Resilience in Knowledge Graph Embeddingscited by 1
  5. 2025Argumentative Ensembling for Robust Recourse under Model Multiplicitycited by 1
  6. 2026Link Prediction or Perdition: the Seeds of Instability in Knowledge Graph Embeddingscited by 1
  7. 2026Towards Foundation Models for Relational Databases with Language Models and Graph Neural Networkscited by 0
  8. 2026SCAIR: Schema-Conditioned Agentic Iterative Reasoning for Enterprise Knowledge Graphscited by 0
  9. 2026Oracle, will I ever learn? A study of prediction convergence and complementarity across link prediction modelscited by 0