CRF-based NER Algorithm

A CRF-based NER Algorithm is a supervised NER algorithm that applies a CRF-based Text Tagging Algorithm (CRF-based Tagger).

• AKA: CRF-based NER, CRF-based NER Tagger.
• Context:
  • It can be implemented in a CRF-based NER System (a CRF-based Sequential Tagging System).
• Example(s):
  • (Tasi & al, 2006).
  • (Minkov & al, 2005).
  • (McCallum & Li, 2003).
• Counter-Example(s):
  • an HMM-based NER Algorithm.
  • an MEMM-based NER Algorithm.
  • a MaxEnt-based NER Algorithm.
• See: SVM-based NER Algorithm, NER Feature.

---

---

References

2007

• (Finkel, 2007) ⇒ Jenny Rose Finkel. (2007). "Named Entity Recognition and the Stanford NER Software." Stanford NLP Lab
  • QUOTE:

	Speed	Discriminative vs. Generative	Normalization
HMM	very fast	generative	local
MEMM	mid-range	discriminative	local
CRF	kinda-slow	discriminative	global

2006

• (Tasi & al, 2006) ⇒ Tzong-han Tsai, Wen-Chi Chou, Shih-Hung Wu, Ting-Yi Sung, Jieh Hsiang, and Wen-Lian Hsu. (2006). "Integrating Linguistic Knowledge into a Conditional Random Field Framework to Identify Biomedical Named Entities." In: Expert Systems with Applications: An International Journal, 30(1). doi:10.1016/j.eswa.2005.09.072

2005

• (Minkov & al, 2005) ⇒ Einat Minkov, Richard C. Wang, and William W. Cohen. (2005). "Extracting personal names from email: applying named entity recognition to informal text." In: Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Language Processing. doi:10.3115/1220575.1220631

2003

• (McCallum & Li, 2003) ⇒ Andrew McCallum, and Wei Li. (2003). "Early Results for Named Entity Recognition with Conditional Random Fields, Feature Induction and Web-Enhanced Lexicons." In: Proceedings of Seventh Conference on Natural Language Learning (CoNLL 2003).

---