pioNER: Datasets and Baselines for Armenian Named Entity Recognition

TL;DR

This paper introduces new Armenian NER datasets, baseline models, and word embeddings, advancing the resources available for Armenian language processing.

Contribution

It provides the first large-scale Armenian NER datasets, trained baseline models, and Armenian-specific word embeddings, facilitating future research.

Findings

Created a 163,000-token Wikipedia-based NER corpus.

Developed a 53,400-token manually annotated news corpus.

Released Armenian GloVe embeddings trained on diverse texts.

Abstract

In this work, we tackle the problem of Armenian named entity recognition, providing silver- and gold-standard datasets as well as establishing baseline results on popular models. We present a 163000-token named entity corpus automatically generated and annotated from Wikipedia, and another 53400-token corpus of news sentences with manual annotation of people, organization and location named entities. The corpora were used to train and evaluate several popular named entity recognition models. Alongside the datasets, we release 50-, 100-, 200-, 300-dimensional GloVe word embeddings trained on a collection of Armenian texts from Wikipedia, news, blogs, and encyclopedia.