{
"title": "DaN+: Danish Nested Named Entities and Lexical Normalization",
"authors": [
{
"name": "Barbara Plank",
"email": "bapl@itu.dk",
"affiliation": "IT University of Copenhagen",
}, {
"name": "Kristian Nørgaard Jensen",
"email": "krnj@itu.dk",
"affiliation": "IT University of Copenhagen"
}, {
"name": "Rob van der Goot",
"email": "robv@itu.dk",
"affiliation": "IT University of Copenhagen"
} ],
"submission_date": "01-04-2021",
"github_link": "https://github.com/bplank/DaNplus",
"paper_link": "https://www.aclweb.org/anthology/2020.coling-main.583.pdf",
"allennlp_version": "1.1",
"datasets": [
{
"name": "DaN+",
"link": "https://github.com/bplank/DaNplus"
}
],
"tags": ["named entity recognition", "named entity detection", "lexical normalization", "domain adaptation", "Danish"]
}
Description:
This paper introduces DAN+, a new multi-domain corpus and annotation guidelines for Danish nested named entities (NEs) and lexical normalization to support research on cross-lingual cross-domain learning for a less-resourced language. We empirically assess three strategies to model the two-layer Named Entity Recognition (NER) task. We compare transfer capabilities from German versus in-language annotation from scratch. We examine language-specific versus multilingual BERT, and study the effect of lexical normalization on NER. Our results show that 1) the most robust strategy is multi-task learning which is rivaled by multi-label decoding, 2) BERT-based NER models are sensitive to domain shifts, and 3) in-language BERT and lexical normalization are the most beneficial on the least canonical data. Our results also show that an out-of-domain setup remains challenging, while performance on news plateaus quickly. This highlights the importance of cross-domain evaluation of cross-lingual transfer.
Project metadata:
Description:
This paper introduces DAN+, a new multi-domain corpus and annotation guidelines for Danish nested named entities (NEs) and lexical normalization to support research on cross-lingual cross-domain learning for a less-resourced language. We empirically assess three strategies to model the two-layer Named Entity Recognition (NER) task. We compare transfer capabilities from German versus in-language annotation from scratch. We examine language-specific versus multilingual BERT, and study the effect of lexical normalization on NER. Our results show that 1) the most robust strategy is multi-task learning which is rivaled by multi-label decoding, 2) BERT-based NER models are sensitive to domain shifts, and 3) in-language BERT and lexical normalization are the most beneficial on the least canonical data. Our results also show that an out-of-domain setup remains challenging, while performance on news plateaus quickly. This highlights the importance of cross-domain evaluation of cross-lingual transfer.