| 1 |
An odd couple – Corpus frequency and look-up frequency: what relationship?
|
|
|
|
In: Slovenščina 2.0: Empirične, aplikativne in interdisciplinarne raziskave, Vol 2, Iss 2, Pp 94-113 (2014) (2014)
|
|
Abstract:
In this paper, we investigate the relationship between log file records and corpus frequency. The study was motivated by practical considerations of how best to keep an already existing corpus-based dictionary updated. Should the next word in the dictionary be the one that follows next on a list of declining corpus frequency? Or the one that users most frequently look up but don’t find? In order to establish manageable criteria, we analysed log files for The Danish Dictionary from 2009 to 2012 and compared the list of most popular words looked up by the users with the frequency of the same words in the corpus underlying The Danish Dictionary. The users’ actual search behaviour was analysed in order to find answers to questions such as these: Are there words which are never looked up? If so, can we say something meaningful about their corpus frequency patterns – do they belong to particular parts of speech, are they particularly frequent or infrequent, could it even be that the pattern is cumulative, in such a way that a particular threshold can be identified? Ultimately, the question is whether it makes sense to use corpus frequency as a criterion for lemma selection.
|
|
Keyword:
corpus frequency; lemma selection; log files; look-up behaviour; P1-1091; Philology. Linguistics; updating dictionaries
|
|
URL: https://doaj.org/article/8ddb1593275e4055af30c8a9da2b16ce
|
|
BASE
|
|
Hide details
|
|
|
|
|
