Prompt-tuning with Attribute Guidance for Low-resource Entity Matching

Lihui Liu; Carl Yang

cs.CLMar 17, 2026

Prompt-tuning with Attribute Guidance for Low-resource Entity Matching

Authors: Lihui Liu, Carl Yang

Organizations: Wayne State University, Detroit, MI, USA · Emory University, Atlanta, GA, USA

Abstract

Entity Matching (EM) is an important task that determines the logical relationship between two entities, such as Same, Different, or Undecidable. Traditional EM approaches rely heavily on supervised learning, which requires large amounts of high-quality labeled data. This labeling process is both time-consuming and costly, limiting practical applicability. As a result, there is a strong need for low-resource EM methods that can perform well with minimal labeled data. Recent prompt-tuning approaches have shown promise for low-resource EM, but they mainly focus on entity-level matching and often overlook critical attribute-level information. In addition, these methods typically lack interpretability and explainability. To address these limitations, this paper introduces PROMPTATTRIB, a comprehensive solution that tackles EM through attribute-level prompt tuning and logical reasoning. PROMPTATTRIB uses both entity-level and attribute-level prompts to incorporate richer contextual information and employs fuzzy logic formulas to infer the final matching label. By explicitly considering attributes, the model gains a deeper understanding of the entities, resulting in more accurate matching. Furthermore, PROMPTATTRIB integrates dropout-based contrastive learning on soft prompts, inspired by SimCSE, which further boosts EM performance. Extensive experiments on real-world datasets demonstrate the effectiveness of PROMPTATTRIB.

Figures & tables

Figure 2: The illustration of prompt-tuning. The blue rectangles in the figure are special prompt tokens, whose parameters are initialized and learnable during prompt-tuning.

Figure 3: The entity level prompt tuning and the attribute level prompt tuning.

Figure 4: Contrastive learning framework. Dropout is applied to the entire input embedding.

Dataset	geo-heter			cameras			computers			ISWC
	F	P	A	F	P	A	F	P	A	F	P	A
SentenceBERT	58.8	53.8	64.8	38.3	21.4	21.3	39.4	35.6	57.5	66.2	68.0	75.5
DeepMatcher	43.8	28.9	72.1	39.2	22.4	23.0	42.8	37.1	58.9	69.1	70.6	76.3
Ditto	3.6	45.5	71.3	42.8	27.3	27.3	44.9	40.3	62.1	72.1	72.7	80.6
PromptEM	78.5	84.5	85.5	35.4	29.4	68.8	49.5	45.2	68.4	76.4	73.6	81.9
PromptAttrib	81.1	85.4	88.9	45.5	45.6	72.2	47.7	49.5	72.1	77.5	79.6	82.5

Table 1: The entity matching performance of different methods. F means F1-score, P means average precision, A means Accuracy.

	F	P	A
BERT	77.2	73.5	83.4
Roberta	79.5	78.2	86.4
Albert	78.7	80.2	86.6
GPT2	64.7	68.4	79.7
Roberta-large	81.1	85.4	88.9
Albert-large	79.9	83.1	86.2

Table 2: The entity matching performance of different language models as backbone. Dataset is geo-heter.

Model	F	P	A
Ditto	42.9	27.3	27.3
PromptEM	42.3	31.7	71.3
PromptAttrib No	40.4	39.7	71.7
PromptAttrib 0.35	45.5	45.6	72.3
PromptAttrib 0.4	37.6	39.0	70.9
PromptAttrib 0.45	39.1	39.5	73.1

Table 3: The entity matching performance with contrastive learning under different dropout ratios. Dataset is cameras.

Entity Resolution

Entity resolution links records that refer to the same real-world entity across datasets despite differences in names, attributes, formats, and languages.

See all 40 papers

Cost-Efficient RAG for Entity Matching with LLMs: A Blocking-based Batch Processing Framework
Feb 5, 2026Chuangtao Ma, Zeyu Zhang, Arijit Khan +2Entity Resolution Retrieval-Augmented Generation
University of Amsterdam Amsterdam, the Netherlands · BIFOLD & TU Berlin Berlin, Germany
Unlocking the Power of Large Language Models for Multi-table Entity Matching
Apr 23, 2026Yingkai Tang, Taoyu Su, Wenyuan Zhang +2Entity Resolution LLM Prompting
Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China · School of Cyber Security, University of Chinese Academy of Sciences, Beijing, China · School of Information Engineering China University of Geosciences Beijing, Beijing, China
Labeling Training Data for Entity Matching Using Large Language Models
Jun 27, 2026Aaron Steiner, Christian BizerLLM-Assisted Annotation Knowledge Distillation
Data and Web Science Group University of Mannheim Mannheim, Germany

Feb 5, 2026J/K move · Enter open · S save

Cost-Efficient RAG for Entity Matching with LLMs: A Blocking-based Batch Processing Framework

Chuangtao Ma, Zeyu Zhang, Arijit Khan, Sebastian Schelter+1

University of Amsterdam Amsterdam, the Netherlands · BIFOLD & TU Berlin Berlin, Germany

Entity Resolution Retrieval-Augmented Generation Graph Retrieval-Augmented Generation

Open paper AskPDF arXiv

Large language models (LLMs) offer a compelling approach to entity matching (EM) in label-scarce settings, yet their matching quality is sensitive to records with ambiguous or sparse attributes due to the absence of external factual grounding. Retrieval-augmented generation (RAG) directly addresses this grounding gap by retrieving and injecting relevant external knowledge. However, naively applying RAG to large-scale EM is expensive due to per-pair retrieval calls over large knowledge graphs (KGs) and the increase in input tokens caused by enriching prompts with retrieved contextual knowledge. To address the above limitation, we introduce a CE-RAG4EM (Cost-Efficient RAG for Entity Matching) framework that reduces overhead through blocking-based batch retrieval and generation. We investigate the benefits of this framework through a systematic analysis of RAG systems for entity matching, focusing on blocking-aware batch optimizations and retrieval granularity. Extensive experiments on nine widely used real-world entity matching benchmark datasets show that CE-RAG4EM under a zero-shot setting achieves comparable or improved matching quality while substantially reducing end-to-end overhead. The results demonstrate that our CE-RAG4EM closes the knowledge grounding gap of the LLM-EM baseline and yields significant improvements in F1 scores on datasets with ambiguous or numerically sparse attributes, while its batch retrieval and inference substantially reduce the overhead and maintain competitive matching quality compared to naive RAG4EM. Our analysis identifies clear performance-overhead trade-offs across blocking-based batch sizes and retrieval granularity, showing how matching quality and computational cost shift with each design choice. These findings offer concrete guidance for building scalable, efficient RAG systems for entity matching that balance accuracy and cost.

Abstract

Entity Resolution

Explore similar work

Cost-Efficient RAG for Entity Matching with LLMs: A Blocking-based Batch Processing Framework

Unlocking the Power of Large Language Models for Multi-table Entity Matching

Labeling Training Data for Entity Matching Using Large Language Models