Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Neural-Symbolic Reasoning over Knowledge Graphs: A Survey from a Query Perspective

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This survey maps knowledge graph reasoning by query type and neural-symbolic method, from single-hop to LLM-integrated reasoning.

desk verdict A well-scoped survey with a sensible query-perspective framing, but the draft is too sloppy — duplicated definitions, a miscategorized TensorLog, and a complex-query section that blurs EFO-1 and tree-formed queries — to be a dependable reference yet. read the letter →

arxiv 2412.10390 v1 pith:N3H7LDOZ submitted 2024-11-30 cs.AI

Lihui Liu , Zihao Wang , Hanghang Tong This is my paper · ORCID
classification cs.AI
keywords knowledgegraphreasoningneural-symboliccomplexlogicalqueriesqueryembeddingslargelanguagemodelscompletionsurvey
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This survey tries to establish that knowledge graph reasoning is best understood through the kinds of queries being asked, and that the methods for answering them fall along a neural-symbolic spectrum. It organizes the field into four query settings: single-hop links, complex logical queries, natural-language questions, and LLM-integrated reasoning. If the organization is right, researchers can use the survey as a map: find the query type, then locate the appropriate symbolic, neural, or hybrid method. The survey also gives formal definitions that separate existential first-order queries from tree-formed queries, which clarifies what different benchmarks and methods actually cover.

What carries the argument

The organizing device is the query-type taxonomy combined with a two-axis classification of symbolic versus neural reasoning. The named formal pair that carries the complex-query section is the EFO-1 query family, existential first-order queries in disjunctive normal form, and the tree-formed (TF) query family, whose answer sets can be built recursively from set operations; the paper uses this distinction to explain why some methods are compositional while others need search or constraint satisfaction. For the LLM section, the machinery is a three-way division of KG-LLM interaction: knowledge graph enhances language model, language model enhances knowledge graph reasoning, and mutual enhancement.

What would settle it

Check a sample of the cited methods against their own source papers using the survey's own category definitions; for instance, read TensorLog's description and ask whether the survey's label of soft symbolic rule based reasoning matches a differentiable deductive database with learned weights. If a substantial fraction of sampled methods land in a different category under the survey's own criteria, the classification is not reliable.

Watch

Extended reading notes

Core claim

The paper's central claim is that a query-centered taxonomy can account for the main lines of knowledge graph reasoning work. For single-hop queries it separates symbolic rule, path, and rule-mining methods from neural-symbolic embeddings and hybrid path or rule learners. For complex logical queries it distinguishes two formal query families, existential first-order (EFO-1) and tree-formed (TF), and argues that embedding methods emphasize set operations while neural-symbolic methods treat queries as constraint-satisfaction searches guided by neural link predictors. For natural-language queries it reviews semantic parsing, retrieval, and embedding approaches in single-turn and multi-turn settings. Finally it claims LLM-KG integration takes three forms: KG helps LLM, LLM helps KG, and mutual enhancement, with the survey as a whole intended as a reference map across all four areas.

Load-bearing premise

The survey's utility rests on the accuracy of its condensed descriptions of individual methods; if enough of those one-line summaries are wrong, the taxonomy can mislead rather than clarify.

Editorial extensions

If this is right

  • A researcher facing a single-hop completion task can pick a method class from the survey's symbolic or neural-symbolic split and know its basic score-function design.
  • For complex queries, the EFO-1 versus tree-form distinction predicts whether a method will model set operations in embedding space or search for variable assignments with a link predictor.
  • The survey's three-way LLM integration scheme gives a shared vocabulary for positioning new systems, such as whether the LLM is used as a translator, an agent, or a joint training target.
  • The formal query definitions imply that many empirical results are scoped to the tree-formed family, so claims about complex query answering should be checked against which query family the benchmark actually uses.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: a reader could use the EFO-1 versus tree-form distinction to predict when box or vector query embeddings degrade, since cyclic query graphs break the compositional set-operation view, a consequence the survey only states obliquely.
  • Editorial inference: the taxonomy invites a stress test where cited methods are checked against their own papers' framing, and the placement of methods like TensorLog as soft symbolic rule reasoning would shift the neural-symbolic boundary if revised.
  • Editorial inference: the LLM-and-KG section opens a fourth direction the survey does not name, using LLMs to build or extend the knowledge graph itself, with graph construction treated as a form of reasoning.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. This manuscript is a survey of knowledge graph reasoning organized by query perspective. It covers single-hop queries, complex logical queries, natural language queries, and the integration of large language models with knowledge graph reasoning, proposing a taxonomy of symbolic, neural, and neural-symbolic methods. It introduces formal definitions for existential first-order (EFO-1) and tree-formed queries and classifies methods according to whether they operate on tree-formed queries, EFO-1 queries, or both. The stated goal is to provide a reliable reference for researchers and practitioners; the paper's utility therefore depends on the accuracy of its formal definitions and its characterizations of cited methods.

Significance. If corrected, the survey would fill a useful niche by organizing knowledge graph reasoning methods around query types and the neural-symbolic distinction, and by incorporating recent formal work on EFO-1 versus tree-formed queries. Its coverage is broad, and it usefully connects classical rule-based reasoning, embedding-based methods, and LLM-based approaches. The paper also gives credit to recent formal characterizations, such as the distinction between EFO-1 and tree-formed query families. However, the current manuscript contains multiple formal and factual errors in load-bearing definitions and method classifications, and because the contribution is a reference survey, these errors directly undermine the central claim of providing a reliable overview. No code or formal proofs are supplied; the contribution is bibliographic and organizational.

major comments (4)
  1. [§2.2, Definitions 1 and 2] Definitions 1 and 2 are textually identical: both define G=(V,E,R) with V as entities, E as relationships, R as triples, and both conclude with "Knowledge graph reasoning: Answer queries by traversing and reasoning over the graph." A definition of knowledge graph reasoning should not be identical to the definition of a knowledge graph. Moreover, the notation is internally inconsistent and propagates into the later formal definitions: R is used for the set of triples in Definition 1, but Definition 3 uses r in E for relations, and Definition 6 uses "a in E" where a is an entity while entities are in V. These errors in the foundational notation weaken the formal taxonomy in Section 4.
  2. [§3.1.2, TensorLog] TensorLog is classified under "Soft symbolic rule based reasoning," but the cited paper [11] is titled "TensorLog: A differentiable deductive database." TensorLog's matrix operations have learnable weights trained by gradient-based optimization, and it is commonly treated as a neural-symbolic system rather than a symbolic one. This placement also contradicts §3.2.2, where Neural LP is correctly described as a generalization of TensorLog within the neural-symbolic category. The taxonomy is therefore internally inconsistent for a method that the survey itself uses as a building block.
  3. [§4.3] The paragraph claiming that "almost all symbolic algorithms search for a proper assignment of variables" and that CQD is beam search on acyclic query graphs, QTO is search with additional backtracking, and FIT generalizes this to cyclic and multi-edge query graphs misrepresents the cited methods. In particular, FIT [76] is a fuzzy-logic method based on query computation trees whose correctness argument addresses EFO-1 queries by expanding existential quantifiers into set operations; it is not simply a backtracking search over variable assignments. The survey itself states in §4.2.2 that non-tree EFO-1 queries cannot be viewed as compositional operations, yet §4.3 treats the EFO-1 and tree-formed query families as interchangeable. This is the load-bearing distinction of the paper's central taxonomy, and as written it will mislead readers about which methods actually handle cyclic or non-tree EFO-1 queries.
  4. [§4.1, EFO-k claim] The text states "so far, there is no specific method targeting EFO-k" while the immediately following citation [75] is titled "EFOk-CQA: Towards Knowledge Graph Complex Query Answering beyond Set Operation." The survey never discusses EFOk-CQA or any other EFO-k method, leaving an unsupported negative claim about the literature. If [75] is intended as a benchmark rather than a method, the sentence should say so; as written, it contradicts the bibliography.
minor comments (5)
  1. [§3.1.1] The text contains the incomplete sentence "Besides rule-based expert system," immediately before the paragraph introducing Prolog.
  2. [§2.1] The text refers to "Fig. 10 in Appendix A" and "Appendices F-A1 and F-A2," but the manuscript contains no appendices; these references should be removed or the appendices supplied.
  3. [Throughout] There are numerous typos and grammatical errors, including "representitive" (§3.1.2), "encludience space" (§3.2.1), "limitions" (§3.2.1), "compliment" for "complement" (§4.1), and "retrain language models" in §6.0.1, which should likely be "train" or "fine-tune."
  4. [§5.2] "Google's Lambda" should be "Google's LaMDA"; the cited reference [56] is the LaMDA paper.
  5. [§4.2.2] The paragraph begins "This part only focuses on methods that are capable of solving queries that are EFO-1 but not tree-form," but the following discussion of LMPNN and Query2Triple does not explain how these methods differ in their treatment of the query graph from the tree-form methods; a clarifying sentence about query-family coverage would help.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the survey's taxonomy is a literature synthesis, and its self-citations are not load-bearing reductions.

full rationale

This paper is a survey rather than a derivation, so there are no predicted quantities fitted from data and no equations whose outputs are defined in terms of their inputs. The central taxonomy in Section 4, distinguishing EFO-1 queries from tree-formed queries, is presented as a synthesis of prior work with definitions given in Definitions 3 through 6 and with citations to [65], [66], and [76]. Two of the survey authors appear among the authors of those cited papers, but the survey does not reduce its classification claim to those citations: the definitions are stated in the text, and the cited works are independently published and externally checkable. No uniqueness theorem from the authors' prior work is invoked to declare the survey's organization forced, and no ansatz is smuggled in solely through citation. The skeptical concern that Section 4 mischaracterizes methods such as CQD, QTO, and FIT is a correctness and utility risk, not a circularity risk: a misleading summary is not equivalent to its inputs by construction. The paper's strongest claim is that it offers a thorough review, and that claim stands or falls on the accuracy of its literature summaries, not on a self-referential derivation. Therefore, no significant circularity is found.

Assumptions & free parameters 0 free parameters · 0 assumptions · 0 invented entities

This is a survey; it introduces no equations, parameters, or entities. The implicit assumption is that the organized taxonomy reflects the field accurately, which is already captured in the weakest_assumption field.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Neural-Symbolic Reasoning over Knowledge Graphs: A Survey from a Query Perspective." pith.science (2026). https://pith.science/paper/N3H7LDOZ

@misc{pith2026241210390,
  author       = {Pith},
  title        = {Pith review of: Neural-Symbolic Reasoning over Knowledge Graphs: A Survey from a Query Perspective},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N3H7LDOZ}},
  note         = {Machine review of arXiv:2412.10390}
}
read the original abstract

Knowledge graph reasoning is pivotal in various domains such as data mining, artificial intelligence, the Web, and social sciences. These knowledge graphs function as comprehensive repositories of human knowledge, facilitating the inference of new information. Traditional symbolic reasoning, despite its strengths, struggles with the challenges posed by incomplete and noisy data within these graphs. In contrast, the rise of Neural Symbolic AI marks a significant advancement, merging the robustness of deep learning with the precision of symbolic reasoning. This integration aims to develop AI systems that are not only highly interpretable and explainable but also versatile, effectively bridging the gap between symbolic and neural methodologies. Additionally, the advent of large language models (LLMs) has opened new frontiers in knowledge graph reasoning, enabling the extraction and synthesis of knowledge in unprecedented ways. This survey offers a thorough review of knowledge graph reasoning, focusing on various query types and the classification of neural symbolic reasoning. Furthermore, it explores the innovative integration of knowledge graph reasoning with large language models, highlighting the potential for groundbreaking advancements. This comprehensive overview is designed to support researchers and practitioners across multiple fields, including data mining, AI, the Web, and social sciences, by providing a detailed understanding of the current landscape and future directions in knowledge graph reasoning.

Figures

Figures reproduced from arXiv: 2412.10390 by the authors.

Figure 1
Figure 1. Survey framework. in the data, making neural reasoning a promising paradigm for var￾ious applications, including question answering, recommendation systems, and semantic understanding. 2.5 Neural Symbolic Reasoning Neural symbolic reasoning represents a fusion of neural network￾based approaches with symbolic reasoning techniques, aiming to leverage the strengths of both paradigms in handling complex rea￾soning tasks… view at source ↗
Figure 2
Figure 2. Rule based expert system. One of the earliest symbolic rule reasoning methods can be track back to 1970s, which is the rule-based expert system [1] as shown in [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Example of Tensorlog. An example of Tensorlog is inferring family relations like “uncle” in [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Three ways to combine LLMs with knowledge graph reasoning. [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Top Ten Challenges Towards Agentic Neural Graph Databases

    cs.AI 2025-01 unverdicted novelty 3.0 of 10

    Agentic Neural Graph Databases are proposed as graph databases with autonomous query construction, neural query execution, and continuous learning, with ten open challenges listed.

Reference graph

Works this paper leans on

89 extracted references · 51 canonical work pages · cited by 1 Pith paper

  1. [76]

    Hang Yin, Zihao Wang, and Yangqiu Song. 2024. Rethinking Complex Queries on Knowledge Graphs with Neural Link Predictors. In The Twelfth International Conference on Learning Representations

    work page 2024

  2. [11]

    William W Cohen. 2016. Tensorlog: A differentiable deductive database. arXiv preprint arXiv:1605.06523 (2016)

    arXiv 2016

  3. [75]

    2024.EFO𝑘 -CQA: Towards Knowledge Graph Complex Query Answering beyond Set Operation

    Hang Yin, Zihao Wang, Weizhi Fei, and Yangqiu Song. 2024.EFO𝑘 -CQA: Towards Knowledge Graph Complex Query Answering beyond Set Operation. (2024)

    work page 2024

  4. [1]

    Ajith Abraham. 2005. Rule-Based expert systems. Handbook of measuring system design (2005)

    2005

  5. [2]

    A Acharya and S Adhikari. 2021. Alexa Conversations: An Extensible Data-driven Approach for Building Task-oriented Dialogue Systems. (2021)

    2021

  6. [3]

    Erik Arakelyan, Daniel Daza, Pasquale Minervini, and Michael Cochez. 2020. Complex query answering with neural link predictors. arXiv preprint arXiv:2011.03459 (2020)

    arXiv 2020

  7. [4]

    Erik Arakelyan, Pasquale Minervini, Daniel Daza, Michael Cochez, and Isabelle Augenstein. 2024. Adapting neural link predictors for data-efficient complex query answering. Advances in Neural Information Processing Systems 36 (2024)

    2024

  8. [5]

    Yushi Bai, Xin Lv, Juanzi Li, and Lei Hou. 2023. Answering Complex Logical Queries on Knowledge Graphs via Query Computation Tree Optimization. In Proceedings of the 40th International Conference on Machine Learning

    2023

Show all 89 references
  1. [6]

    Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Jason Weston, and Ok- sana Yakhnenko. 2013. Translating embeddings for modeling multi-relational data. Advances in neural information processing systems 26 (2013)

    2013

  2. [7]

    C Buck and J Bulian. 2017. Ask the Right Questions: Active Question Reformula- tion with Reinforcement Learning. (2017)

    2017

  3. [8]

    Stefano Ceri, Georg Gottlob, Letizia Tanca, et al. 1989. What you always wanted to know about Datalog(and never dared to ask). IEEE transactions on knowledge and data engineering 1, 1 (1989), 146–166

    1989

  4. [9]

    Xuelu Chen, Ziniu Hu, and Yizhou Sun. 2022. Fuzzy logic based logical query an- swering on knowledge graphs. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 36. 3939–3948

    2022

  5. [10]

    William F Clocksin and Christopher S Mellish. 2003. Programming in PROLOG. Springer Science & Business Media

    2003

  6. [12]

    R Das, S Dhuliawala, and M Zaheer. 2017. Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning

    2017

  7. [13]

    Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt, and Jonathan Larson. 2024. From Local to Global: A Graph RAG Approach to Query-Focused Summarization.arXiv preprint arXiv:2404.16130 (2024)

    2024 arXiv

  8. [14]

    A Elgohary, D Peskov, and J Boyd-Graber. 2019. Can You Unpack That? Learning to Rewrite Questions-in-Context. Association for Computational Linguistics

    2019

  9. [15]

    Katrin Erk. 2009. Representing Words as Regions in Vector Space. InProceedings of the Thirteenth Conference on Computational Natural Language Learning (Boulder, Colorado) (CoNLL ’09). Association for Computational Linguistics, USA, 57–65

    2009

  10. [16]

    A Fader, L Zettlemoyer, and O Etzioni. 2014. Open Question Answering over Curated and Extracted Knowledge Bases (KDD ’14). Association for Computing Machinery

    2014

  11. [17]

    Luis Galárraga, Christina Teflioudi, Katja Hose, and Fabian M Suchanek. 2015. Fast rule mining in ontological knowledge bases with AMIE+. The VLDB Journal 24, 6 (2015), 707–730

    2015

  12. [18]

    Luis Antonio Galárraga, Christina Teflioudi, Katja Hose, and Fabian Suchanek

  13. [19]

    Matt Gardner, Partha Talukdar, Jayant Krishnamurthy, and Tom Mitchell. 2014. Incorporating Vector Space Similarity in Random Walk Inference over Knowl- edge Bases. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), Alessandro Mos...

    2014 doi

  14. [20]

    D Guo, D Tang, N Duan, M Zhou, and J Yin. 2018. Dialog-to-Action: Con- versational Question Answering Over a Large-Scale Knowledge Base. Curran Associates, Inc

    2018

  15. [21]

    Will Hamilton, Payal Bajaj, Marinka Zitnik, Dan Jurafsky, and Jure Leskovec

  16. [22]

    Shizhu He, Kang Liu, Guoliang Ji, and Jun Zhao. 2015. Learning to Represent Knowledge Graphs with Gaussian Embedding. In Proceedings of the 24th ACM International on Conference on Information and Knowledge Management (Mel- bourne, Australia) (CIKM ’15). Association for Computi...

    2015

  17. [23]

    Vinh Thinh Ho, Daria Stepanova, Mohamed H Gad-Elrab, Evgeny Kharlamov, and Gerhard Weikum. 2018. Rule learning from knowledge graphs guided by embedding models. In The Semantic Web–ISWC 2018: 17th International Semantic Web Conference. Springer, 72–90

    2018

  18. [24]

    E Kacupaj, J Plepi, K Singh, and H Thakkar. [n. d.]. Conversational Question Answering over Knowledge Graphs with Transformer and Graph Attention Networks. Association for Computational Linguistics

  19. [25]

    Endri Kacupaj, Kuldeep Singh, Maria Maleshkova, and Jens Lehmann. 2022. Contrastive Representation Learning for Conversational Question Answering over Knowledge Graphs. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management (CIKM ’22) . A...

    2022

  20. [26]

    Magdalena Kaiser, Rishiraj Saha Roy, and Gerhard Weikum. 2021. Reinforcement learning from reformulations in conversational question answering over knowl- edge graphs. In Proceedings of the 44th international ACM SIGIR conference on research and development in information retr...

    2021

  21. [27]

    Jiho Kim, Yeonsu Kwon, Yohan Jo, and Edward Choi. 2023. KG-GPT: A general framework for reasoning on knowledge graphs using large language models. arXiv preprint arXiv:2310.11220 (2023)

    2023 arXiv

  22. [28]

    Ni Lao, Tom Mitchell, and William Cohen. 2011. Random walk inference and learning in a large scale knowledge base. In Proceedings of the 2011 conference on empirical methods in natural language processing . 529–539

    2011

  23. [29]

    Leonid Libkin. 2004. Elements of finite model theory . Vol. 41. Springer

    2004

  24. [30]

    X Lin and R Socher. [n. d.]. Multi-Hop Knowledge Graph Reasoning with Reward Shaping. In EMNLP 2018

    2018

  25. [31]

    Yankai Lin, Zhiyuan Liu, Maosong Sun, Yang Liu, and Xuan Zhu. 2015. Learning entity and relation embeddings for knowledge graph completion. In Proceedings of the AAAI conference on artificial intelligence , Vol. 29

    2015

  26. [32]

    Lihui Liu, Yuzhong Chen, Mahashweta Das, Hao Yang, and Hanghang Tong. 2023. Knowledge Graph Question Answering with Ambiguous Query. In Proceedings of the ACM Web Conference 2023

    2023

  27. [33]

    Lihui Liu, Boxin Du, Heng Ji, ChengXiang Zhai, and Hanghang Tong. 2021. Neural-Answering Logical Queries on Knowledge Graphs. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD ’21)

    2021

  28. [34]

    Lihui Liu, Boxin Du, Jiejun Xu, Yinglong Xia, and Hanghang Tong. 2022. Joint Knowledge Graph Completion and Question Answering. InProceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (Washington DC, USA) (KDD ’22). Association for Computing Mach...

    2022

  29. [35]

    Lihui Liu, Blaine Hill, Boxin Du, Fei Wang, and Hanghang Tong. 2024. Con- versational Question Answering with Language Models Generated Reformu- lations over Knowledge Graph. In Findings of the Association for Computa- tional Linguistics: ACL 2024 , Lun-Wei Ku, Andre Martins, ...

    2024 doi

  30. [36]

    Lihui Liu, Zihao Wang, Ruizhong Qiu, Yikun Ban, Eunice Chan, Yangqiu Song, Jingrui He, and Hanghang Tong. 2024. Logic Query of Thoughts: Guiding Large Language Models to Answer Complex Logic Queries with Knowledge Graphs. arXiv:2404.04264 [cs.IR] https://arxiv.org/abs/2404.04264

    2024 arXiv

  31. [37]

    David Marker. 2006. Model theory: an introduction . Vol. 217. Springer Science & Business Media. Neural-Symbolic Reasoning over Knowledge Graphs: A Survey from a Query Perspective Conference’17, July 2017, Washington, DC, USA

    2006

  32. [38]

    Deborah L McGuinness, Frank Van Harmelen, et al. 2004. OWL web ontology language overview. W3C recommendation 10, 10 (2004), 2004

    2004

  33. [39]

    Alexander Miller, Adam Fisch, Jesse Dodge, Amir-Hossein Karimi, Antoine Bor- des, and Jason Weston. 2016. Key-Value Memory Networks for Directly Reading Documents. arXiv:1606.03126 [cs.CL]

    2016 arXiv

  34. [40]

    T Misu, K Georgila, A Leuski, and D Traum. 2012. Reinforcement Learning of Question-Answering Dialogue Policies for Virtual Museum Guides. Association for Computational Linguistics

    2012

  35. [41]

    Maximilian Nickel, Lorenzo Rosasco, and Tomaso Poggio. 2016. Holographic embeddings of knowledge graphs. In Proceedings of the AAAI conference on artificial intelligence, Vol. 30

    2016

  36. [42]

    Maximilian Nickel, Volker Tresp, Hans-Peter Kriegel, et al. 2011. A three-way model for collective learning on multi-relational data.. In Icml, Vol. 11. 3104482– 3104584

    2011

  37. [43]

    Pouya Ghiasnezhad Omran, Kewen Wang, and Zhe Wang. 2018. Scalable Rule Learning via Learning Representation.. In IJCAI. 2149–2155

    2018

  38. [44]

    Meng Qu, Junkun Chen, Louis-Pascal Xhonneux, Yoshua Bengio, and Jian Tang

  39. [45]

    A Radford, J Wu, and R Child. 2018. Language Models are Unsupervised Multitask Learners. (2018)

    2018

  40. [46]

    Hongyu Ren, Weihua Hu, and Jure Leskovec. 2020. Query2box: Reasoning over knowledge graphs in vector space using box embeddings. arXiv preprint arXiv:2002.05969 (2020)

    2020 arXiv

  41. [47]

    Hongyu Ren and Jure Leskovec. 2020. Beta embeddings for multi-hop logical reasoning in knowledge graphs. Advances in Neural Information Processing Systems 33 (2020), 19716–19726

    2020

  42. [48]

    Matthew Richardson and Pedro Domingos. 2006. Markov logic networks.Machine learning 62 (2006), 107–136

    2006

  43. [49]

    Apoorv Saxena, Aditay Tripathi, and Partha Talukdar. 2020. Improving multi-hop question answering over knowledge graphs using knowledge base embeddings. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 4498–4507

    2020

  44. [50]

    Weijia Shi, Sewon Min, Michihiro Yasunaga, Minjoon Seo, Rich James, Mike Lewis, Luke Zettlemoyer, and Wen-tau Yih. 2023. Replug: Retrieval-augmented black-box language models. arXiv preprint arXiv:2301.12652 (2023)

    2023 arXiv

  45. [51]

    Haitian Sun, Andrew Arnold, Tania Bedrax Weiss, Fernando Pereira, and William W Cohen. 2020. Faithful embeddings for knowledge base queries. Ad- vances in Neural Information Processing Systems 33 (2020), 22505–22516

    2020

  46. [52]

    Haitian Sun, Tania Bedrax-Weiss, and William W. Cohen. 2019. PullNet: Open Domain Question Answering with Iterative Retrieval on Knowledge Bases and Text. arXiv:1904.09537 [cs.CL]

    2019 arXiv

  47. [53]

    Jiashuo Sun, Chengjin Xu, Lumingyuan Tang, Saizhuo Wang, Chen Lin, Yeyun Gong, Heung-Yeung Shum, and Jian Guo. 2023. Think-on-graph: Deep and responsible reasoning of large language model with knowledge graph. arXiv preprint arXiv:2307.07697 (2023)

    2023 arXiv

  48. [54]

    Zhiqing Sun, Zhi-Hong Deng, Jian-Yun Nie, and Jian Tang. 2019. Rotate: Knowl- edge graph embedding by relational rotation in complex space. arXiv preprint arXiv:1902.10197 (2019)

    2019 arXiv

  49. [55]

    Komal Teru, Etienne Denis, and Will Hamilton. 2020. Inductive relation prediction by subgraph reasoning. In International Conference on Machine Learning . PMLR, 9448–9457

    2020

  50. [56]

    Romal Thoppilan, Daniel De Freitas, and Jamie Hall. 2022. LaMDA: Language Models for Dialog Applications. arXiv. https://doi.org/10.48550/ARXIV.2201. 08239

    2022 doi

  51. [57]

    Théo Trouillon, Johannes Welbl, Sebastian Riedel, Éric Gaussier, and Guillaume Bouchard. 2016. Complex embeddings for simple link prediction. In International conference on machine learning . PMLR, 2071–2080

    2016

  52. [58]

    S Vakulenko, S Longpre, Z Tu, and R Anantha. [n. d.]. Question Rewriting for Conversational Question Answering (WSDM ’21). Association for Computing Machinery

  53. [59]

    Luke Vilnis, Xiang Li, Shikhar Murty, and Andrew McCallum. 2018. Probabilistic Embedding of Knowledge Graphs with Box Lattice Measures. In ACL

    2018

  54. [60]

    Luke Vilnis and Andrew McCallum. 2015. Word Representations via Gaussian Embedding. arXiv:1412.6623 [cs.CL]

    2015 arXiv

  55. [61]

    Hongwei Wang, Hongyu Ren, and Jure Leskovec. 2021. Relational message passing for knowledge graph completion. InProceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining . 1697–1707

    2021

  56. [62]

    Xiaozhi Wang, Tianyu Gao, Zhaocheng Zhu, Zhengyan Zhang, Zhiyuan Liu, Juanzi Li, and Jian Tang. 2021. KEPLER: A unified model for knowledge embed- ding and pre-trained language representation. Transactions of the Association for Computational Linguistics 9 (2021), 176–194

    2021

  57. [63]

    Zihao Wang, Weizhi Fei, Hang Yin, Yangqiu Song, Ginny Wong, and Simon See

  58. [64]

    Zihao Wang, Yangqiu Song, Ginny Wong, and Simon See. 2023. Logical Message Passing Networks with One-hop Inference on Atomic Formulas. In The Eleventh International Conference on Learning Representations

    2023

  59. [65]

    Zihao Wang, Hang Yin, and Yangqiu Song. 2022. Benchmarking the Combinatorial Generalizability of Complex Query Answering on Knowledge Graphs.Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks 1 (NeurIPS Datasets and Benchmarks 2021) (2022)

    2022

  60. [66]

    Zihao Wang, Hang Yin, and Yangqiu Song. 2022. Logical Queries on Knowledge Graphs: Emerging Interface of Incomplete Relational Data. (2022)

    2022

  61. [67]

    Zhen Wang, Jianwen Zhang, Jianlin Feng, and Zheng Chen. 2014. Knowledge graph embedding by translating on hyperplanes. In Proceedings of the AAAI conference on artificial intelligence , Vol. 28

    2014

  62. [68]

    Wenhan Xiong, Thien Hoang, and William Yang Wang. 2017. Deeppath: A reinforcement learning method for knowledge graph reasoning. arXiv preprint arXiv:1707.06690 (2017)

    2017 arXiv

  63. [69]

    Yao Xu, Shizhu He, Cunguang Wang, Li Cai, Kang Liu, and Jun Zhao. 2023. Query2Triple: Unified Query Encoding for Answering Diverse Complex Queries over Knowledge Graphs. arXiv preprint arXiv:2310.11246 (2023)

    2023

  64. [70]

    Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 2014. Em- bedding entities and relations for learning and inference in knowledge bases. arXiv preprint arXiv:1412.6575 (2014)

    2014 arXiv

  65. [71]

    Dong Yang, Peijun Qing, Yang Li, Haonan Lu, and Xiaodong Lin. 2022. GammaE: Gamma Embeddings for Logical Queries on Knowledge Graphs. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing . 745–760

    2022

  66. [72]

    Fan Yang, Zhilin Yang, and William W Cohen. 2017. Differentiable learning of logical rules for knowledge base completion. CoRR, abs/1702.08367 (2017)

    2017 arXiv

  67. [73]

    Michihiro Yasunaga, Hongyu Ren, Antoine Bosselut, Percy Liang, and Jure Leskovec. 2021. QA-GNN: Reasoning with language models and knowledge graphs for question answering. arXiv preprint arXiv:2104.06378 (2021)

    2021 arXiv

  68. [74]

    Xi Ye, Semih Yavuz, Kazuma Hashimoto, Yingbo Zhou, and Caiming Xiong

  69. [77]

    Donghan Yu, Chenguang Zhu, Yiming Yang, and Michael Zeng. 2020. JAKET: Joint Pre-training of Knowledge Graph and Language Understanding. arXiv:2010.00796 [cs.CL]

    2020 arXiv

  70. [78]

    Chongzhi Zhang, Zhiping Peng, Junhao Zheng, and Qianli Ma. 2024. Conditional logical message passing transformer for complex query answering. InProceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 4119–4130

    2024

  71. [79]

    Qixuan Zhang, Xinyi Weng, Guangyou Zhou, Yi Zhang, and Jimmy Xiangji Huang. 2022. ARL: An adaptive reinforcement learning framework for complex question answering over knowledge base. Information Processing and Manage- ment 59, 3 (2022), 102933. https://doi.org/10.1016/j.ipm.2...

    2022

  72. [80]

    Shuai Zhang, Yi Tay, Lina Yao, and Qi Liu. 2019. Quaternion knowledge graph embeddings. Advances in neural information processing systems 32 (2019)

    2019

  73. [81]

    Xikun Zhang, Antoine Bosselut, Michihiro Yasunaga, Hongyu Ren, Percy Liang, Christopher D Manning, and Jure Leskovec. 2022. Greaselm: Graph reasoning en- hanced language models for question answering. arXiv preprint arXiv:2201.08860 (2022)

    2022 arXiv

  74. [82]

    Zhanqiu Zhang, Jie Wang, Jiajun Chen, Shuiwang Ji, and Feng Wu. 2021. Cone: Cone embeddings for multi-hop reasoning over knowledge graphs. Advances in Neural Information Processing Systems 34 (2021), 19172–19183

    2021

  75. [83]

    Li Zhou, Jianfeng Gao, Di Li, and Heung-Yeung Shum. 2020. The Design and Implementation of XiaoIce, an Empathetic Social Chatbot. Computa- tional Linguistics 46, 1 (03 2020), 53–93. https://doi.org/10.1162/coli_a_00368 arXiv:https://direct.mit.edu/coli/article-pdf/46/1/53/1847...

    2020 doi

  76. [84]

    Zhaocheng Zhu, Mikhail Galkin, Zuobai Zhang, and Jian Tang. 2022. Neural- symbolic models for logical queries on knowledge graphs. In International con- ference on machine learning . PMLR, 27454–27478

    2022

  77. [2013]

    In Proceedings of the 22nd international conference on World Wide Web

    AMIE: association rule mining under incomplete evidence in ontological knowledge bases. In Proceedings of the 22nd international conference on World Wide Web. 413–422

  78. [2018]

    Advances in neural information processing systems 31 (2018)

    Embedding logical queries on knowledge graphs. Advances in neural information processing systems 31 (2018)

    2018

  79. [2020]

    arXiv preprint arXiv:2010.04029 (2020)

    Rnnlogic: Learning logic rules for reasoning on knowledge graphs. arXiv preprint arXiv:2010.04029 (2020)

    2020 arXiv

  80. [2021]

    In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL)

    RnG-KBQA: Generation Augmented Iterative Ranking for Knowledge Base Question Answering. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL)

  81. [2023]

    In Findings of the Association for Computa- tional Linguistics: ACL 2023

    Wasserstein-Fisher-Rao Embedding: Logical Query Embeddings with Local Comparison and Global Transport. In Findings of the Association for Computa- tional Linguistics: ACL 2023. 13679–13696

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.