Neural Architectures for Named Entity Recognition
Нейронные архитектуры для распознавания именованных сущностей
2016-01-01
SCID: 54.1/dzz3366r
Discuss with AI
CRFLSTMNamed Entity RecognitionNeural ArchitecturesSequence labeling
Figures from the paper
Abstract (AI)
Guillaume Lample, Miguel Ballesteros, Sandeep Subramanian, Kazuya Kawakami, Chris Dyer. Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2016.
Key Findings
1
A neural network architecture for named entity recognition (NER) is presented that does not rely on gazetteers or manually engineered features.
2
End-to-end training of the neural architectures yields competitive or superior results on standard NER benchmarks compared to feature-engineered systems.
3
The approach generalizes across languages by learning subword and word patterns, reducing dependence on language-specific resources.
4
The proposed models combine character-level and word-level representations to improve NER performance.
Research Object
Named entity recognition systems (neural architectures for NER)
Research Subject
Design and evaluation of neural network architectures for performing named entity recognition
Publication Details
Publication Date
2016-01-01
Journal
Publisher
ISSN
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest