Re-ranking via local embeddings : a use case with permutation-based indexing and the nSimplex projection
- Submitting institution
-
University of St Andrews
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 267881924
- Type
- D - Journal article
- DOI
-
10.1016/j.is.2020.101506
- Title of journal
- Information Systems
- Article number
- 101506
- First page
- -
- Volume
- 95
- Issue
- -
- ISSN
- 0306-4379
- Open access status
- Deposit exception
- Month of publication
- February
- Year of publication
- 2020
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
5
- Research group(s)
-
B - Systems
- Citation count
- 0
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The MI File is a state-of-the-art approximate search technique developed by Amato and Rabitti at ISTI-CNR Pisa, one of the world’s leading similarity search groups. It has excellent recall, but relatively poor precision. This article shows how our n-simplex projection can be used to cheaply re-rank its outcome, combining the best features of both mechanisms. The journal article is a significantly extended version of a conference paper which won the best paper award at SISAP18 in Lima.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
