# Magneto: Combining Small and Large Language Models for Schema Matching

arXiv: https://arxiv.org/abs/2412.08194 · HTML: https://mlsanity.com/paper/2412.08194/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 82 references that are in the map.
- First posted on arXiv 11 Dec 2024.

## Builds on (most cited first, 22 of 22)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [Text Embeddings by Weakly-Supervised Contrastive Pre-training](https://mlsanity.com/paper/2212.03533/) (2022)
- [MPNet: Masked and Permuted Pre-training for Language Understanding](https://mlsanity.com/paper/2004.09297/) (2020)
- [In Defense of the Triplet Loss for Person Re-Identification](https://mlsanity.com/paper/1703.07737/) (2017)
- [TaBERT: Pretraining for Joint Understanding of Textual and Tabular Data](https://mlsanity.com/paper/2005.08314/) (2020)
- [Long-context LLMs Struggle with Long In-context Learning](https://mlsanity.com/paper/2404.02060/) (2024)
- [Can Foundation Models Wrangle Your Data?](https://mlsanity.com/paper/2205.09911/) (2022)
- [Large Language Models(LLMs) on Tabular Data: Prediction, Generation, and Understanding -- A Survey](https://mlsanity.com/paper/2402.17944/) (2024)
- [Semantics-aware Dataset Discovery from Data Lakes with Contextualized Column-based Representation Learning](https://mlsanity.com/paper/2210.01922/) (2022)
- [Valentine: Evaluating Matching Techniques for Dataset Discovery](https://mlsanity.com/paper/2010.07386/) (2020)
- [OLaLa: Ontology Matching with Large Language Models](https://mlsanity.com/paper/2311.03837/) (2023)
- [SANTOS: Relationship-based Semantic Table Union Search](https://mlsanity.com/paper/2209.13589/) (2022)
- [Schema Matching with Large Language Models: an Experimental Study](https://mlsanity.com/paper/2407.11852/) (2024)
- [DeepJoin: Joinable Table Discovery with Pre-trained Language Models](https://mlsanity.com/paper/2212.07588/) (2022)
- [ArcheType: A Novel Framework for Open-Source Column Type Annotation using Large Language Models](https://mlsanity.com/paper/2310.18208/) (2023)
- [Pylon: Semantic Table Union Search in Data Lakes](https://mlsanity.com/paper/2301.04901/) (2023)
- [KcMF: A Knowledge-compliant Framework for Schema and Entity Matching with Fine-tuning-free LLMs](https://mlsanity.com/paper/2410.12480/) (2024)
- [Sampling Methods for Inner Product Sketching](https://mlsanity.com/paper/2309.16157/) (2023)

## Built on by (most cited first, 17 of 17)

- [A Survey of Data Agents: Emerging Paradigm or Overstated Hype?](https://mlsanity.com/paper/2510.23587/) (2025)
- [Can LLMs Clean Up Your Mess? A Survey of Application-Ready Data Preparation with LLMs](https://mlsanity.com/paper/2601.17058/) (2026)
- [Empowering Tabular Data Preparation with Language Models: Why and How?](https://mlsanity.com/paper/2508.01556/) (2025)
- [Interactive Data Harmonization with LLM Agents: Opportunities and Challenges](https://mlsanity.com/paper/2502.07132/) (2025)
- [Automatic End-to-End Data Integration using Large Language Models](https://mlsanity.com/paper/2603.10547/) (2026)
- [Database Normalization via Dual-LLM Self-Refinement](https://mlsanity.com/paper/2508.17693/) (2025)
- [RACT: Retrieval Augmented Column-Table Learning and Prediction for Multi-Table Schema Matching](https://mlsanity.com/paper/2606.07843/) (2026)
- [MaDI-Bench: An End-to-End Data Integration Benchmark](https://mlsanity.com/paper/2606.30371/) (2026)
- [Conceptual Schema Inference for Tabular Datasets using Large Language Models](https://mlsanity.com/paper/2509.04632/) (2025)
- [CONE: Embeddings for Complex Numerical Data Preserving Unit and Variable Semantics](https://mlsanity.com/paper/2603.04741/) (2026)
- [BDI-Kit Demo: A Toolkit for Programmable and Conversational Data Harmonization](https://mlsanity.com/paper/2604.06405/) (2026)
- [SemStruct: Contextualizing Semantic Embeddings with Structural Information for Schema Matching](https://mlsanity.com/paper/2605.30729/) (2026)
- [Single and Multi Truth Data Fusion using Large Language Models](https://mlsanity.com/paper/2606.28062/) (2026)
- [Alignment-Guided Largest Table Overlap Size Estimation](https://mlsanity.com/paper/2607.03049/) (2026)
- [AgentMap: Joint Equivalence and Subsumption Discovery for Ontology Matching](https://mlsanity.com/paper/2607.27130/) (2026)
- [Discovering Relationships in Data Lakes Using Large Language Models: An Industrial Case](https://mlsanity.com/paper/2608.26750/) (2026)
- [Surprising Effectiveness of Self-Demonstrations in Enhancing Schema-Ontology Mapping with LLMs](https://mlsanity.com/paper/2609.13776/) (2026)

## Abstract (arXiv)

Recent advances in language models opened new opportunities to address complex schema matching tasks. Schema matching approaches have been proposed that demonstrate the usefulness of language models, but they have also uncovered important limitations: Small language models (SLMs) require training data (which can be both expensive and challenging to obtain), and large language models (LLMs) often incur high computational costs and must deal with constraints imposed by context windows. We present Magneto, a cost-effective and accurate solution for schema matching that combines the advantages of SLMs and LLMs to address their limitations. By structuring the schema matching pipeline in two phases, retrieval and reranking, Magneto can use computationally efficient SLM-based strategies to derive candidate matches which can then be reranked by LLMs, thus making it possible to reduce runtime without compromising matching accuracy. We propose a self-supervised approach to fine-tune SLMs which uses LLMs to generate syntactically diverse training data, and prompting strategies that are effective for reranking. We also introduce a new benchmark, developed in collaboration with domain experts, which includes real biomedical datasets and presents new challenges to schema matching methods. Through a detailed experimental evaluation, using both our new and existing benchmarks, we show that Magneto is scalable and attains high accuracy for datasets from different domains.