Named Entity Recognition in Industrial Tables using Tabular Language Models

TL;DR

This paper explores applying transformer-based models to industrial table data for Named Entity Recognition, addressing challenges like limited labeled data with a novel augmentation strategy, and demonstrating the importance of tabular structure for model performance.

Contribution

It introduces a domain-specific data augmentation method and highlights the significance of tabular inductive bias in transformer models for industrial NER tasks.

Findings

Table transformers outperform baselines in industrial NER.

Data augmentation significantly improves low-resource performance.

Tabular structure is crucial for model convergence.

Abstract

Specialized transformer-based models for encoding tabular data have gained interest in academia. Although tabular data is omnipresent in industry, applications of table transformers are still missing. In this paper, we study how these models can be applied to an industrial Named Entity Recognition (NER) problem where the entities are mentioned in tabular-structured spreadsheets. The highly technical nature of spreadsheets as well as the lack of labeled data present major challenges for fine-tuning transformer-based models. Therefore, we develop a dedicated table data augmentation strategy based on available domain-specific knowledge graphs. We show that this boosts performance in our low-resource scenario considerably. Further, we investigate the benefits of tabular structure as inductive bias compared to tables as linearized sequences. Our experiments confirm that a table transformer outperforms other baselines and that its tabular inductive bias is vital for convergence of transformer-based models.