Papers › Fine-tuning Large Language Models for Entity Matching

Fine-tuning Large Language Models for Entity Matching

12 Sep 2024arXiv:2409.08185archive 2025-07-28

Aaron Steiner, Ralph Peeters, Christian Bizer

Generative large language models (LLMs) are a promising alternative to pre-trained language models for entity matching due to their high zero-shot performance and ability to generalize to unseen entities. Existing research on using LLMs for entity matching has focused on prompt engineering and in-context learning. This paper explores the potential of fine-tuning LLMs for entity matching. We analyze fine-tuning along two dimensions: 1) the representation of training examples, where we experiment with adding different types of LLM-generated explanations to the training set, and 2) the selection and generation of training examples using LLMs. In addition to the matching performance on the source dataset, we investigate how fine-tuning affects the models ability to generalize to other in-domain datasets as well as across topical domains. Our experiments show that fine-tuning significantly improves the performance of the smaller models while the results for the larger models are mixed. Fine-tuning also improves the generalization to in-domain datasets while hurting cross-domain transfer. We show that adding structured explanations to the training set has a positive impact on the performance of three out of four LLMs, while the proposed example selection and generation methods, only improve the performance of Llama 3.1 8B while decreasing the performance of GPT-4o-mini.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

wbsg-uni-mannheim/tailormatch officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Data IntegrationEntity ResolutionIn-Context LearningPrompt Engineering

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Entity Resolution Abt-Buy gpt-4o-mini-2024-07-18_fine_tuned F1 (%) 94.09 #3 of 16 Archive leaderboard report
Entity Resolution Abt-Buy gpt-4o-2024-08-06 F1 (%) 92.20 #4 of 16 Archive leaderboard report
Entity Resolution Abt-Buy gpt-4o-mini-2024-07-18 F1 (%) 87.68 #8 of 16 Archive leaderboard report
Entity Resolution Abt-Buy Meta-Llama-3.1-8B-Instruct_fine_tuned F1 (%) 87.34 #9 of 16 Archive leaderboard report
Entity Resolution Abt-Buy Meta-Llama-3.1-70B-Instruct F1 (%) 79.12 #12 of 16 Archive leaderboard report
Entity Resolution Abt-Buy Meta-Llama-3.1-8B-Instruct F1 (%) 56.57 #15 of 16 Archive leaderboard report
Entity Resolution Amazon-Google gpt-4o-mini-2024-07-18_fine_tuned F1 (%) 80.25 #2 of 17 Archive leaderboard report
Entity Resolution Amazon-Google gpt-4o-2024-08-06 F1 (%) 63.45 #12 of 17 Archive leaderboard report
Entity Resolution Amazon-Google gpt-4o-mini-2024-07-18 F1 (%) 59.20 #13 of 17 Archive leaderboard report
Entity Resolution Amazon-Google Meta-Llama-3.1-70B-Instruct F1 (%) 51.44 #15 of 17 Archive leaderboard report
Entity Resolution Amazon-Google Meta-Llama-3.1-8B-Instruct_fine_tuned F1 (%) 50.00 #16 of 17 Archive leaderboard report
Entity Resolution Amazon-Google Meta-Llama-3.1-8B-Instruct F1 (%) 49.16 #17 of 17 Archive leaderboard report
Entity Resolution WDC Products gpt-4o-2024-08-06_fine_tuned_wdc_small F1 (%) 87.07 #1 of 1 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium gpt-4o-2024-08-06_fine_tuned_wdc_small F1 (%) 87.10 #2 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium gpt-4o-mini-2024-07-18_structured_explanations F1 (%) 84.38 #3 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium gpt-4o-mini-2024-07-18 F1 (%) 81.61 #4 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium Llama3.1_70B_structured_explanations F1 (%) 76.70 #6 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium Llama3.1_70B F1 (%) 75.20 #7 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium Llama3.1_8B_error-based_example_selection F1 (%) 74.37 #8 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium Llama3.1_8B_structured_explanations F1 (%) 74.13 #9 of 13 Archive leaderboard report
Entity Resolution WDC Products-80%cc-seen-medium Llama3.1_8B F1 (%) 53.36 #13 of 13 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLLaMALabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSETSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections