OpenNER 1.0: Standardized Open-Access Named Entity Recognition Datasets in 50+ Languages

2024-12-12

arXiv.org

AI Summary

Existing NER research is hindered by the lack of standardized, openly accessible, and structurally homogeneous multilingual and multi-ontology datasets. To address this, we introduce the first open-source, standardized NER resource repository covering 51 languages and 34 datasets. Our resource unifies annotation formats (CoNLL), entity type naming conventions, and data schemas, and establishes a cross-ontology entity type mapping specification. This work represents the first systematic integration and standardization of multilingual and multi-ontology NER data, resolving longstanding bottlenecks—including format heterogeneity, non-comparable entity taxonomies, and restricted data access. We build a unified benchmark using mBERT, XLM-R, and InfoXLM, providing ready-to-use data splits and comprehensive baseline performance reports. The resource substantially lowers the barrier to multilingual NER experimentation and has already enabled multiple model reproductions and evaluations, thereby enhancing comparability, reproducibility, and progress in multilingual and multi-ontology NER research.

Technology Category

• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Knowledge Representation and Reasoning: Ontologies
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

We present OpenNER 1.0, a standardized collection of openly available named entity recognition (NER) datasets. OpenNER contains 34 datasets spanning 51 languages, annotated in varying named entity ontologies. We correct annotation format issues, standardize the original datasets into a uniform representation, map entity type names to be more consistent across corpora, and provide the collection in a structure that enables research in multilingual and multi-ontology NER. We provide baseline models using three pretrained multilingual language models to compare the performance of recent models and facilitate future research in NER.

Problem

• Standardizing multilingual NER datasets for consistency
• Providing baseline results using diverse language models
• Addressing performance gaps in LLMs for NER tasks

Innovation

• Standardized multilingual NER datasets
• Uniform entity type representation
• Baseline results with multiple models