RECIPE-TKG: From Sparse History to Structured Reasoning for LLM-based Temporal Knowledge Graph Completion

ACL ARR 2025 July Submission1196 Authors

Abstract: Temporal Knowledge Graphs (TKGs) represent dynamic facts as timestamped relations between entities. TKG completion involves forecasting missing or future links, requiring models to reason over time-evolving structure. While LLMs show promise for this task, existing approaches often overemphasize supervised fine-tuning and struggle particularly when historical evidence is limited or missing. We introduce RECIPE-TKG, a lightweight and data-efficient framework designed to improve accuracy and generalization in settings with sparse historical context. It combines (1) rule-based multi-hop retrieval for structurally diverse history, (2) contrastive fine-tuning of lightweight adapters to encode relational semantics, and (3) test-time semantic filtering to iteratively refine generations based on embedding similarity. Experiments on four TKG benchmarks show that RECIPE-TKG outperforms previous LLM-based approaches, achieving up to 22.4\% relative improvement in Hits@10. Moreover, our proposed framework produces more semantically coherent predictions, even for the samples with limited historical context.

Paper Type: Long

Research Area: Machine Learning for NLP

Research Area Keywords: Machine Learning for NLP, NLP Applications, Efficient/Low-Resource Methods for NLP, Information Retrieval and Text Mining, Language Modeling

Contribution Types: NLP engineering experiment, Approaches to low-resource settings, Approaches low compute settings-efficiency, Publicly available software and/or pre-trained models

Languages Studied: English

Previous URL: https://openreview.net/forum?id=s7sq3KdQnG

Reassignment Request Area Chair: Yes, I want a different area chair for our submission

Reassignment Request Reviewers: Yes, I want a different set of reviewers

Justification For Not Keeping Action Editor Or Reviewers: We respectfully request new reviewers for this submission based on several concerns with the previous review process: Despite our comprehensive rebuttal, we received no engagement from reviewers despite multiple follow-up inquiries. Several reviewer questions addressed information already present in our appendices, suggesting incomplete review of our materials. Some citation requests appeared inconsistent with standard practice in temporal knowledge graph completion literature, as the papers in question are not commonly referenced in recent ACL/EMNLP publications in this domain. Additionally, several reviews arrived after the deadline and contained questions readily answerable from the main text and appendix, raising concerns about review thoroughness. We have substantially revised the paper to address all concerns and believe a fresh evaluation by reviewers with expertise in LLM-based knowledge graph reasoning would best serve the assessment of our work.

A1 Limitations Section: This paper has a limitations section.

A2 Potential Risks: N/A

B Use Or Create Scientific Artifacts: Yes

B1 Cite Creators Of Artifacts: Yes

B1 Elaboration: References

B2 Discuss The License For Artifacts: Yes

B2 Elaboration: License and Ethics

B3 Artifact Use Consistent With Intended Use: N/A

B4 Data Contains Personally Identifying Info Or Offensive Content: N/A

B5 Documentation Of Artifacts: N/A

B6 Statistics For Data: N/A

C Computational Experiments: Yes

C1 Model Size And Budget: Yes

C1 Elaboration: Appendix B.4

C2 Experimental Setup And Hyperparameters: Yes

C2 Elaboration: Section 5.1 & Appendix B.4

C3 Descriptive Statistics: Yes

C3 Elaboration: Section 5 and Appendix

C4 Parameters For Packages: N/A

D Human Subjects Including Annotators: No

D1 Instructions Given To Participants: N/A

D2 Recruitment And Payment: N/A

D3 Data Consent: N/A

D4 Ethics Review Board Approval: N/A

D5 Characteristics Of Annotators: N/A

E Ai Assistants In Research Or Writing: Yes

E1 Information About Use Of Ai Assistants: Yes

E1 Elaboration: Appendix J

Author Submission Checklist: yes

Submission Number: 1196