Global Predictor Feature

A global predictor feature is a predictor feature that represents a Global Relationship.

• AKA: Global Feature.
• Context:
  • It can make use of information from the entire object (e.g. an entire Discourse).
  • …
• Counter-Example(s):
  • a Local Predictor Feature.
  • a Long-Range Predictor Feature.
• See: Context Feature, Long-Distance Relationship.

---

---

References

• http://scholar.google.com/scholar?q=%22global+feature%22+training+%22local+feature%22

2006

• (Krishnan & Manning, 2006) ⇒ V. Krishnan, and Christopher D. Manning. (2006). “An Effective Two-stage Model for Exploiting Non-local Dependencies in Named Entity Recognition.” In: Proceedings of the 21st International Conference on Computational Linguistics and the 44th Annual Meeting of the Association for Computational Linguistics. (ACL-2006) doi:10.3115/1220175.1220316

---