Few-shot Name Entity Recognition on StackOverflow

TL;DR

This paper introduces RoBERTa+MAML, a meta-learning based few-shot NER method tailored for StackOverflow data, achieving notable F1 score improvements with domain-specific enhancements.

Contribution

The paper presents a novel few-shot NER approach combining RoBERTa and MAML, specifically designed for StackOverflow's challenging domain with limited labeled data.

Findings

Achieved 5% F1 score improvement over baseline

Effective domain-specific phrase processing enhances results

Demonstrated viability of meta-learning for domain-specific NER

Abstract

StackOverflow, with its vast question repository and limited labeled examples, raise an annotation challenge for us. We address this gap by proposing RoBERTa+MAML, a few-shot named entity recognition (NER) method leveraging meta-learning. Our approach, evaluated on the StackOverflow NER corpus (27 entity types), achieves a 5% F1 score improvement over the baseline. We improved the results further domain-specific phrase processing enhance results.