Supervised Sequence-Member Classification Task

AKA: Supervised Labeling, Supervised Sequence Tagging.
Context:
- It is a type-of Supervised Structured-Input Classification Task.
- It can be solved by a Supervised Sequence-Member Classification System (that implements a supervised sequence-member classification algorithm).
- It can be used to solve a Supervised Chunking Task (e.g. with BIO tags).
Example(s):
- a Supervised Text Tagging Task, such as supervised POS tagging.
- …
Counter-Example(s):
- a Supervised Sequence Classification Task.
- a Supervised Graph Tagging Task.
See: Supervised Classification Task.

References

(Graves, 2012) ⇒ Alex Graves. (2012). “Supervised Sequence Labelling.” In: Supervised Sequence Labelling with Recurrent Neural Networks, pp. 5-13 . Springer Berlin Heidelberg,

(Mejer & Crammer, 2010) ⇒ Avihai Mejer, and Koby Crammer. (2010). “Confidence in Structured-prediction Using Confidence-weighted Models.” In: Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing (EMNLP 2010).
- QUOTE: In the sequence labeling setting, instances [math]\displaystyle{ x }[/math] belong to a general input space [math]\displaystyle{ \mathcal{X} }[/math] and conceptually are composed of a finite number [math]\displaystyle{ n }[/math] of components, such as words of a sentence. The number of components [math]\displaystyle{ n = |x| }[/math] varies between instances. Each part of a instance is labelled from a finite set [math]\displaystyle{ \mathcal{Y} }[/math], [math]\displaystyle{ |\mathcal{Y}| = K }[/math]. That is, a labeling of an entire instance belongs to the product set [math]\displaystyle{ y \in \mathcal{Y} × \mathcal{Y} ... \mathcal{Y} }[/math] ([math]\displaystyle{ n }[/math] times).