Output details
11 - Computer Science and Informatics
University College London
Return to search
Output 0 of 0 in the submission
Output title
A new rank correlation coefficient for information retrieval.
Type
E - Conference contribution
Name of conference/published proceedings
SIGIR
Volume number
-
Issue number
-
First page of article
587
ISSN of proceedings
-
Year of publication
2008
Number of additional authors
2
Additional information
<17> Originality: First paper to propose rank correlation statistics that are top heavy. Significance: Most researchers in information retrieval use rank correlation statistics and reach conclusions based on these statistics. In information retrieval, usually the top end of a ranking is more important than the bottom end of a ranking; however, this fact has been ignored by all rank correlation statistics before this one, causing the researchers to reach invalid conclusions. Rigour: A nice theoretical approach to devise a new rank correlation statistic that has a mathematical interpretation, validated via extensive experiments.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Citation count
43
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-