Hybrid knowledge reasoning over knowledge hypergraph: Inductive, deductive, and abductive

Abstract

Traditional knowledge reasoning methods, which are predominantly reliant on static rules and structured data, often struggle to adapt to the ambiguity and dynamic evolution of real-world scenarios. To overcome these limitations, this study proposes a novel reasoning framework based on a three-layered knowledge hypergraph. Core innovation lies in the synergy of inductive, deductive, and abductive reasoning mechanisms to enhance both reliability and interpretability. Specifically, hypergraph-based inductive reasoning extracts robust evolutionary patterns by mining the historical subgraph structures. Deductive reasoning ensures transparency by constructing tree-shaped inference paths, whereas abductive reasoning establishes causal traceability by forming evidence chains from historical contexts. Experimental evaluations on the integrated crisis early warning system (ICEWS) dataset demonstrate that the proposed approach significantly outperforms existing methods in terms of accuracy and interpretability, thereby offering a scalable solution for complex event analysis.

Keywords

Abductive reasoning / Deductive reasoning / Inductive reasoning / Knowledge hypergraphs / Knowledge reasoning

CRediT authorship contribution statement

Ling Tian: Conceptualization, Methodology, Formal analysis, Visualization, Writing―original draft, Supervision, Project administration, Funding acquisition. Lei Gao: Methodology, Investigation, Writing―original draft, Formal analysis. Ben Zhang: Methodology, Investigation, Writing―original draft, Formal analysis. Xiao Liu: Methodology, Data curation, Validation, Writing―review & editing. Yi-Nong Shi: Writing―review & editing. Hui Gao: Methodology, Conceptualization, Formal analysis.

Declaration of competing interest

The authors declare the following personal relationships which may be considered as potential competing interests: Ling Tian is the committee member for Journal of Electronic Science and Technology and was not involved in the editorial review or the decision to publishing this article. Other authors declare that there are no conflicts of interests.

Acknowledgment

This work was supported by the National Natural Science Foundation of China under Grant No. 62376055.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jnlest.2026.100361.

References

[1] Z.-W. Tao, X.-C. Chen, Z. Jin, et al., EVIT: event-oriented instruction tuning for event reasoning, in: Proc. of the Findings of the Association for Computational Linguistics, 2024, pp. 8966-8979,Bangkok, Thailand.

[2] X. Liu, Y. He, W.-X. Tai, et al., Exploring the chameleon effect of contextual dynamics in temporal knowledge graph for event prediction, Tsinghua Sci. Technol., 3 0 (1) (2025) pp. 433-455.

[3] Z.-W. Tao, Z. Jin, Y.-F. Zhang, et al., A comprehensive evaluation on event reasoning of large language models, in: Proc. of the 39th AAAI Conf. on Artificial Intelligence, 2025, pp. 25273-25281,Philadelphia, USA.

[4] C.-C. Zhu, M.-H. Chen, C.-J. Fan, et al., Learning from history: modeling temporal knowledge graphs with sequential copy-generation networks, in: Proc. of the 35th AAAI Conf. on Artificial Intelligence, Virtual Event, 2021, pp. 4732-4740.

[5] Y.-J. Li, S.-L. Sun, J. Zhao, TiRGN: time-guided recurrent graph network with local-global historical patterns for temporal knowledge graph reasoning, in: Proc. of the 31st Intl. Joint Conf. on Artificial Intelligence, 2022, pp. 2152-2158,Vienna, Austria.

[6] H. Dong, P.-Y. Wang, M. Xiao, et al., Temporal inductive path neural network for temporal knowledge graph reasoning, Artif. Intell. 329 (2024) 104085.

[7] S.-H. Xiong, Y. Yang, A. Payani, et al., TEILP: time prediction over knowledge graphs via logical reasoning, in: Proc. of the 38th AAAI Conf. on Artificial Intelligence, 2024, pp. 16112-16119,Vancouver, Canada.

[8] M.-Q. Chen, Y.-B. Ma, K.-T. Song, et al., Improving large language models in event relation logical prediction, in: Proc. of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2024, pp. 9451-9478,Bangkok, Thailand.

[9] W. Chen, H.-Y. Wan, Y.-T. Wu, et al., Local-global history-aware contrastive learning for temporal knowledge graph reasoning, in: Proc. of the 40th IEEE Intl. Conf. on Data Engineering, 2024, pp. 733-746,Utrecht, The Netherlands.

[10] K. Liang, L.-Y. Meng, M. Liu, et al., Learn from relational correlations and periodic events for temporal knowledge graph reasoning, in: Proc. of the 46th Intl. ACM SIGIR Conf. on Research and Development in Information Retrieval, 2023, pp. 1559-1568,Taipei, China.

[11] W.-H. Ding, H.-H. Lin, B. Li, et al., Generalizing goal-conditioned reinforcement learning with variational causal reasoning, in: Proc. of the 36th Intl. Conf. on Neural Information Processing Systems, 2022, pp. 26532-26548,New Orleans, USA.

[12] Z.-P. Cai, Z.-B. Xiong, H.-H. Xu, et al., Generative adversarial networks: a survey toward private and secure applications, ACM Comput. Surv., 5 4 (6) (2022) p. 132.

[13] L.A. Galárraga, C. Teflioudi, K. Hose, et al., AMIE: association rule mining under incomplete evidence in ontological knowledge bases, in: Proc. of the 22nd Intl. Conf. on World Wide Web, 2013, pp. 413-422,Rio de Janeiro, Brazil.

[14] L. Galárraga, C. Teflioudi, K. Hose, et al., Fast rule mining in ontological knowledge bases with AMIE+, VLDB J, 2 4 (6) (2015) pp. 707-730.

[15] F. Yang, Z.-L. Yang, W.W. Cohen, Differentiable learning of logical rules for knowledge base reasoning, in: Proc. of the 31st Intl. Conf. on Neural Information Processing Systems, Long Beach, USA, 2017, pp. 2316-2325.

[16] A. Sadeghian, M. Armandpour, P. Ding, et al., DRUM: end-to-end differentiable rule mining on knowledge graphs, in: Proc. of the 33rd Intl. Conf. on Neural Information Processing Systems, 2019, pp. 15321-15331,Vancouver, Canada.

[17] T. Rocktäschel, S. Riedel, End-to-end differentiable proving, in: Proc. of the 31st Intl. Conf. on Neural Information Processing Systems, Long Beach, USA, 2017, pp. 3791-3803.

[18] A. Aamodt, E. Plaza, Case-based reasoning: foundational issues, methodological variations, and system approaches, AI Commun., 7 (1) (1994) pp. 39-59.

[19] D. Leake, D. Crandall, On bringing case-based reasoning methodology to deep learning, in: Proc. of the 28th Intl. Conf. on Case-based Reasoning, Research and Development, Salamanca, Spain, 2020, pp. 343-348.

[20] D. Leake, X.-M. Ye, D.J. Crandall, Supporting case-based reasoning with neural networks: an illustration for case adaptation, in: Proc. of the AAAI Spring Symposium on Combining Machine Learning and Knowledge Engineering, 2021,Palo Alto, USA.

[21] Z.-L. Yang, X.-Y. Du, E. Cambria, et al., End-to-end case-based reasoning for commonsense knowledge base completion, in: Proc. of the 17th Conf. of the European Chapter of the Association for Computational Linguistics, Dubrovnik, Croatia, 2023, pp. 3509-3522.

[22] B.-S. Yang, W.-T. Yih, X.-D. He, et al., Embedding entities and relations for learning and inference in knowledge bases, in: Proc. of the 3rd Intl. Conf. on Learning Representations, 2015,San Diego, USA.

[23] T. Trouillon, J. Welbl, S. Riedel, et al., Complex embeddings for simple link prediction, in: Proc. of the 33rd Intl. Conf. on Machine Learning, 2016, pp. 2071-2080,New York, USA.

[24] I. Balazevic, C. Allen, T. Hospedales, TuckER: tensor factorization for knowledge graph completion, in: Proc. of the Conf. on Empirical Methods in Natural Language Proc. of the 9th Intl. Joint Conf. on Natural Language Processing, 2019, pp. 5185-5194,Hong Kong, China.

[25] T. Dettmers, P. Minervini, P. Stenetorp, et al., Convolutional 2D knowledge graph embeddings, in: Proc. of the 32nd AAAI Conf. on Artificial Intelligence, 2018, pp. 1811-1818,New Orleans, USA.

[26] M. Schlichtkrull, T.N. Kipf, P. Bloem, et al., Modeling relational data with graph convolutional networks, in: Proc. of the 15th Intl. Conf. on the Semantic Web, 2018, pp. 593-607,Heraklion, Greece.

[27] W.-H. Xiong, M. Yu, S.-Y. Chang, et al., One-shot relational learning for knowledge graphs, in: Proc. of the Conf. on Empirical Methods in Natural Language Processing, 2018, pp. 1980-1990,Brussels, Belgium.

[28] L.-H. Luo, Y.-F. Li, G. Haffari, et al., Normalizing flow-based neural process for few-shot knowledge graph completion, in: Proc. of the 46th Intl. ACM SIGIR Conf. on Research and Development in Information Retrieval, 2023, pp. 900-910,Taipei, China.

[29] A.W. Burks, Peirce's theory of abduction, Philos. Sci., 1 3 (4) (1946) pp. 301-306.

[30] F.E. Reilly, Charles Peirce's Theory of Scientific Method, Fordham University Press, New York, USA, (2018).

[31] E. Boschee, J. Lautenschlager, S. O'Brien, et al., ICEWS coded event data, [Online]. Available, https://doi.org/10.7910/DVN/28075 (2015).