Definition Extraction Algorithm
A Definition Extraction Algorithm is an information extraction algorithm that is a definition creation algorithm.
- Context:
- It can be implemented by a Definition Extraction System (that solves a definition extraction task).
- It can be supported by a Definitional Sentence Classification Algorithm and/or a Definitional Sentence Retrieval Algorithm.
- It can support a Definitional Question Answering Algorithm.
- See: Rule-based Definition Extraction Algorithm.
References
2019
- (Spala et al., 2019) ⇒ Sasha Spala, Nicholas A. Miller, Yiming Yang, Franck Dernoncourt, and Carl Dockhorn. (2019). “DEFT: A Corpus for Definition Extraction in Free- and Semi-structured Text.” In: Proceedings of the 13th Linguistic Annotation Workshop.
- QUOTE: ... Most related work on definition extraction has relied on the idea that text|definitions can be captured by common “definitor” verb phrases like “means”, “refers to”, and “is”. Early work in the field incorporated rule-based methods that extracted sentences that met this narrow standard (JL Clavens, 2001; Cui and Chua, 2004, 2005; Fahmi and Bouma, 2006; Zhang and Jiang, 2009). While predictable and easily applied, these models subsequently failed to extract sentences that lack these explicit markers. In an effort to expand on the type of phrases used to extract definitions, Cui et al. (2007) used soft pattern matching in a modified HMM (PHMM). More recent work from Espinosa Anke and Schockaert (2018) makes use of a neural approach, which reached state-of-the-art performance on the word class lattices (WCL) datasets (Navigli et al., 2010). Even so, these methods require both term and definition to appear in the same sentence and for terms to appear before definitions.
2013
- (Boella & Di Caro, 2013) ⇒ Guido Boella, and Luigi Di Caro. (2013). “Extracting Definitions and Hypernym Relations Relying on Syntactic Dependencies and Support Vector Machines.” In: Proceedings of the 51st annual meeting of the association for computational linguistics (ACL-2013).