The most-frequent-sense and the pre- dominant domain sense play an impor- tant role in the debate on word-sense- disambiguation. This discussion is, how-.
In this paper, we argue that current sense-tagged corpora neglect rare senses and contexts and, as a result, do not represent a good corpus for training and ...
This paper argues that current sense-tagged corpora neglect rare senses and contexts and, as a result, do not represent a good corpus for training and ...
Dec 1, 2013 · In this paper, we argue that current sense-tagged corpora neglect rare senses and contexts and, as a result, do not represent a good corpus for ...
Missing: quest | Show results with:quest
(PDF) DutchSemCor: Targeting the ideal sense-tagged corpus
www.researchgate.net › publication › 23...
PDF | On Jan 1, 2012, Piek Vossen and others published DutchSemCor: Targeting the ideal sense-tagged corpus | Find, read and cite all the research you need ...
Missing: quest | Show results with:quest
the lemmas got 25 examples for each sense. A small but significant proportion of senses is not well represented in the corpus even after Internet search.
Missing: quest | Show results with:quest
In this paper, we will define quality criteria for sense-tagged corpora and will describe a novel method for building a large- scale sense-tagged Dutch corpus ...
Abstract. Word Sense Disambiguation (WSD) systems require large sense-tagged corpora along with lexical databases to reach satisfactory results.
Missing: quest | Show results with:quest
DutchSemCor: Targeting the ideal sense-tagged corpus ; Publisher, European Language Resources Association (ELRA) ; Pages, 584-589 ; ISBN (Print), 9782951740877.
Missing: quest | Show results with:quest
DutchSemCor: Targeting the ideal sense-tagged corpus ... Word Sense Disambiguation (WSD) systems require large sense-tagged corpora along with lexical databases ...