Deep Entity Matching with Pre-Trained Language Models

Abstract

We present Ditto, a novel entity matching system based on pre-trained Transformer-based language models. We fine-tune and cast EM as a sequence-pair classification problem to leverage such models with a simple architecture. Our experiments show that a straightforward application of language models such as BERT, DistilBERT, or RoBERTa pre-trained on large text corpora already significantly improves the matching quality and outperforms previous state-of-the-art (SOTA), by up to 29% of F1 score on benchmark datasets. We also developed three optimization techniques to further improve Ditto's matching capability. Ditto allows domain knowledge to be injected by highlighting important pieces of input information that may be of interest when making matching decisions. Ditto also summarizes strings that are too long so that only the essential information is retained and used for EM. Finally, Ditto adapts a SOTA technique on data augmentation for text to EM to augment the training data with (difficult) examples. This way, Ditto is forced to learn "harder" to improve the model's matching capability. The optimizations we developed further boost the performance of Ditto by up to 9.8%. Perhaps more surprisingly, we establish that Ditto can achieve the previous SOTA results with at most half the number of labeled data. Finally, we demonstrate Ditto's effectiveness on a real-world large-scale EM task. On matching two company datasets consisting of 789K and 412K records, Ditto achieves a high F1 score of 96.5%.

To appear in VLDB 2021

References in corpus (6)

• BioBERT: a pre-trained biomedical language representation model for biomedical text mining
• KGAT: Knowledge Graph Attention Network for Recommendation
• Deep Entity Matching with Pre-Trained Language Models
• Neural Natural Language Inference Models Enhanced with External Knowledge
• Snippext: Semi-supervised Opinion Mining with Augmented Data
• A Comprehensive Benchmark Framework for Active Learning Methods in Entity Matching

Cited by in corpus (24)

• Deep Entity Matching with Pre-Trained Language Models
• Neural Networks for Entity Matching: A Survey
• Construction of Knowledge Graphs: State and Challenges
• Annotating Columns with Pre-trained Language Models
• OLaLa: Ontology Matching with Large Language Models
• Supervised Contrastive Learning for Product Matching
• LargeEA: Aligning Entities for Large-scale Knowledge Graphs
• Deep Indexed Active Learning for Matching Heterogeneous Entity Representations
• OAG-Bench: A Human-Curated Benchmark for Academic Graph Mining
• Consistent and Flexible Selectivity Estimation for High-Dimensional Data
• R2D2: Reducing Redundancy and Duplication in Data Lakes
• AdapterEM: Pre-trained Language Model Adaptation for Generalized Entity Matching using Adapter-tuning
• "The Data Says Otherwise"-Towards Automated Fact-checking and Communication of Data Claims
• The Battleship Approach to the Low Resource Entity Matching Problem
• Entity Matching by Pool-based Active Learning
• Cross-Language Learning for Entity Matching
• Preliminary Guidelines For Combining Data Integration and Visual Data Analysis
• Multilingual Transformers for Product Matching -- Experiments and a New Benchmark in Polish
• Leveraging Large Language Models for Generating Labeled Mineral Site Record Linkage Data
• GXJoin: Generalized Cell Transformations for Explainable Joinability
• PLM4NDV: Minimizing Data Access for Number of Distinct Values Estimation with Pre-trained Language Models
• TabulaX: Leveraging Large Language Models for Multi-Class Table Transformations
• Unlocking the Power of Large Language Models for Multi-table Entity Matching
• Graph-based Active Learning for Entity Cluster Repair