Labeling Heuristic

A Labeling Heuristic is a heuristic that can convert an unlabeled example to a labeled example (a weakly labeled example).

• AKA: Labeling Pattern.
• Context:
  • It can be used in a Heuristic Labeling Algorithm (such as a Self-Supervised Learning Algorithm).
• See: Labeling Task, Pattern, Self-Supervised Learning Task, Weakly Labeled Dataset.

---

---

References

2012

• (Melli, 2012) ⇒ Gabor Melli. (2012). “Identifying Untyped Relation Mentions in a Corpus given an Ontology.” In: Workshop Proceedings of TextGraphs-7: Graph-based Methods for Natural Language Processing (TextGraphs-7).
  • QUOTE: To overcome the lack of annotated data, we propose a labelling heuristic based on information extracted from the ontology.

2008

• (Banko & Etzioni, 2008) ⇒ Michele Banko, and Oren Etzioni. (2008). “The Tradeoffs Between Open and Traditional Relation Extraction.” In: Proceedings of the 46th Annual Meeting of the Association for Computational Linguistics (ACL 2008).
  • QUOTE: As with O-NB, O-CRF’s training process is a self-supervised. O-CRF applies a handful of relation-independent heuristics to the PennTreebank and obtains a set of labeled examples in the form of relational tuples. The heuristics were designed to capture dependencies typically obtained via syntactic parsing and semantic role labelling.

2003

• (Pinto et al., 2003) ⇒ David Pinto, Andrew McCallum, Xing Wei, and W. Bruce Croft. (2003). “Table Extraction Using Conditional Random Fields.” In: Proceedings of the 26th annual international ACM SIGIR conference on Research and development in information retrieval (SIGIR 2003). doi:10.1145/860435.860479
  • QUOTE: … A simple heuristic program was run to make an educated guess for a label for each line. Human judgment was then employed to correct mistakes in the heuristic labeling. …

---