Please use this identifier to cite or link to this item:
Title: From n-gram to skipgram to concgram
Authors: Cheng, W 
Greaves, C
Warren, M 
Issue Date: 2006
Source: International journal of corpus linguistics, 2006, v. 11, no. 4, p. 411-433
Abstract: Uncovering the extent of word associations and how they are manifested has been an important area of study in corpus linguistics since the 1960s (Sinclair et al. 1970). This paper defines and describes a new way of categorising word association, the concgram, which constitutes all of the permutations of constituency and positional variation generated by the association of two or more words. Concgrams are identified without prior input from the user (other than to set the size of the span) employing a fully automated search that reveals all of the word association patterns that exist in a corpus. This study argues that concgrams represent more fully word associations in a corpus. Most concgrams seem to be non-contiguous, and show both constituency (AB, ACB) and positional (AB, BA) variations. Further studies of concgrams will help in the task of uncovering the full extent of the idiom principle (Sinclair 1987).
Keywords: Concgram
Contiguous and non-contiguous word associations
Constituency and positional variations
Corpus linguistics
Publisher: John Benjamins Publishing Company
Journal: International journal of corpus linguistics 
ISSN: 1384-6655
1569-9811 (E-ISSN)
DOI: 10.1075/ijcl.11.4.04che
Rights: International Journal of Corpus Linguistics Copyright © 2006 John Benjamins Publishing Company.
The publisher should be contacted for permission to re-use or reprint the materials in any form.
Appears in Collections:Journal/Magazine Article

Files in This Item:
File Description SizeFormat 
concgram13.pdfPre-published version1.41 MBAdobe PDFView/Open
View full-text via PolyU eLinks SFX Query
Show full item record
PIRA download icon_1.1View/Download Full Text


Last Week
Last month
Citations as of Sep 6, 2020

Page view(s)

Last Week
Last month
Citations as of Jan 31, 2021


Citations as of Jan 31, 2021

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.