Bidirectional Transformer Encoder-based NER Algorithm

A Bidirectional Transformer Encoder-based NER Algorithm is a neural NER algorithm that employs a Bidirectional Transformer Encoder (to solve am NER task).

• Context:
  • It can (typically) analyze text data in both directions (left-to-right and right-to-left), enabling the extraction of rich contextual embeddings.
  • It can be based on pre-training text corpora and fine-tuned for specific NER tasks.
  • It can handle languages with complex syntactic structures because it can understand context holistically.
  • ...
• Example(s):
  • BERT-MRC (Li, Feng, et al., 2019), which applies a Bidirectional Transformer for question-answering frameworks to the NER task.
  • GLiNER (Zaratiana et al., 2023), a compact model that leverages a Bidirectional Transformer Encoder for efficient and effective named entity recognition across various domains and languages.
  • ...
• Counter-Example(s):
  • LLM-based NER Algorithm.
  • Traditional NER Algorithm.
  • Unsupervised Named Entity Recognition Algorithms that do not use Bidirectional Transformer Encoders for entity recognition.
• See: Transformer Model, Transfer Learning, Contextual Embedding.

---

---

References

2023

• (Zaratiana et al., 2023) ⇒ Urchade Zaratiana, Nadi Tomeh, Pierre Holat, and Thierry Charnois. (2023). “Gliner: Generalist Model for Named Entity Recognition Using Bidirectional Transformer.” arXiv preprint arXiv:2311.08526
  • ABSTRACT: Named Entity Recognition (NER) is essential in various Natural Language Processing (NLP) applications. Traditional NER models are effective but limited to a set of predefined entity types. In contrast, Large Language Models (LLMs) can extract arbitrary entities through natural language instructions, offering greater flexibility. However, their size and cost, particularly for those accessed via APIs like ChatGPT, make them impractical in resource-limited scenarios. In this paper, we introduce a compact NER model trained to identify any type of entity. Leveraging a bidirectional transformer encoder, our model, GLiNER, facilitates parallel entity extraction, an advantage over the slow sequential token generation of LLMs. Through comprehensive testing, GLiNER demonstrate strong performance, outperforming both ChatGPT and fine-tuned LLMs in zero-shot evaluations on various NER benchmarks.

2019

• (Li et al., 2019) ⇒ Xiaoya Li, Jingrong Feng, Yuxian Meng, Qinghong Han, Fei Wu, and Jiwei Li. (2019). “A Unified MRC Framework for Named Entity Recognition.” arXiv preprint arXiv:1910.11476
  • ABSTRACT: The task of named entity recognition (NER) is normally divided into nested NER and flat NER depending on whether named entities are nested or not. Models are usually separately developed for the two tasks, since sequence labeling models, the most widely used backbone for flat NER, are only able to assign a single label to a particular token, which is unsuitable for nested NER where a token may be assigned several labels.

    In this paper, we propose a unified framework that is capable of handling both flat and nested NER tasks. Instead of treating the task of NER as a sequence labeling problem, we propose to formulate it as a machine reading comprehension (MRC) task. For example, extracting entities with the \textsc{per} label is formalized as extracting answer spans to the question "{\it which person is mentioned in the text?}". This formulation naturally tackles the entity overlapping issue in nested NER: the extraction of two overlapping entities for different categories requires answering two independent questions. Additionally, since the query encodes informative prior knowledge, this strategy facilitates the process of entity extraction, leading to better performances for not only nested NER, but flat NER.

    We conduct experiments on both {\em nested} and {\em flat} NER datasets. Experimental results demonstrate the effectiveness of the proposed formulation. We are able to achieve vast amount of performance boost over current SOTA models on nested NER datasets, i.e., +1.28, +2.55, +5.44, +6.37, respectively on ACE04, ACE05, GENIA and KBP17, along with SOTA results on flat NER datasets, i.e.,+0.24, +1.95, +0.21, +1.49 respectively on English CoNLL 2003, English OntoNotes 5.0, Chinese MSRA, Chinese OntoNotes 4.0.

---