Output details
11 - Computer Science and Informatics
University of Brighton
Return to search
Output 0 of 0 in the submission
Output title
Discrete vs. continuous rating scales for language evaluation in NLP
Type
E - Conference contribution
DOI
-
Name of conference/published proceedings
Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics
Volume number
-
Issue number
-
First page of article
230
ISSN of proceedings
-
Year of publication
2011
URL
-
Number of additional authors
1
Additional information
<22>
This paper presents a systematic evaluation of discrete and continuous rating scales used in human quality assessments of computer-generated language. Such studies are common in psychology and related fields. This paper presents the first detailed study of this type for NLP. Continuous rating scales had not been used for this purpose in NLP before and are shown to be viable for use in language evaluation, offering distinct advantages over discrete scales. ACL is a leading conference in the field, being A* rated in the Australian ERA/CORE list, and had an acceptance rate of 25% in 2011.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Citation count
1
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
