Sequence Member Tagging Structure

A sequence member tagging structure is a classification structure that can by applied to a sequence tagging task (to classify each sequence element).

• AKA: String Member Labeling Model.
• Context:
  • It can be used by a Sequence Tagging System (to solve a Sequence Tagging Task (that converts a Sequence to a Tagged Sequence).
  • It can be produced by a Supervised Sequence Tagging Task/Sequence Tagging Model Creation Task, such as:
  • It can be used as a Chunking Model, using BIO Labels.
• Example(s):
  • a Finite-State Sequence Tagging Function.
  • a Heuristic Sequence Tagging Function.
  • a CRF-based Sequence Tagging Function.
  • an MEMM-based Sequence Tagging Function.
  • an HMM-based Sequence Tagging Function.
  • a Maximum-Entropy Function.
  • an SVM-based Sequence Tagging Function.
  • a Text Tagging Function, such as a POS Tagging Function.
  • a DNA Tagging Function.
  • a Stochastic Memoizer.
  • …
• Counter-Example(s):
  • a Chunking Function.
  • a Tree Tagging Function.
  • a Lattice Tagging Function.
  • a Text Tokenization Function.
  • a Sequence Segment Classifier, such as a Named Entity Mention Classifier.
  • a Sequence Classifier, such as a Text Classifier.
  • a Sentence Parser.
  • a Graph Node Classifier.
• See: Data Stream Mining, IOB Tag Set.

---

---

References

2009

• (Ye et al., 2009) ⇒ Nan Ye, Wee Sun Lee, Hai Leong Chieu, and Dan Wu. (2009). “Conditional Random Fields with High-Order Features for Sequence Labeling.” In: Advances in Neural Information Processing Systems 22 (NIPS 2009)

2005

• (Wallach, 2005) ⇒ Hanna M. Wallach. (2005). “Conditional Random Fields." Literature Survey Webpage.
  • Conditional random fields (CRFs) are a probabilistic framework for labeling and segmenting structured data, such as sequences, trees and lattices.

---