2018
DOI: 10.1111/cobi.13117
|View full text |Cite
|
Sign up to set email alerts
|

Using machine learning to advance synthesis and use of conservation and environmental evidence

Abstract: Article impact statement: Machine learning optimizes processes of systematic evidence synthesis and improves its utility for evidence‐based conservation.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
63
0

Year Published

2018
2018
2021
2021

Publication Types

Select...
8
1

Relationship

1
8

Authors

Journals

citations
Cited by 95 publications
(70 citation statements)
references
References 8 publications
(10 reference statements)
0
63
0
Order By: Relevance
“…We also included all references from the bibliographies of the two previous systematic reviews on CO-BSI in Africa and Asia (25,26). Endnote was used to remove duplicates, and a final deduplicated data set was uploaded to an online systematic review tool for abstract and full-text screening (91).…”
Section: Methodsmentioning
confidence: 99%
“…We also included all references from the bibliographies of the two previous systematic reviews on CO-BSI in Africa and Asia (25,26). Endnote was used to remove duplicates, and a final deduplicated data set was uploaded to an online systematic review tool for abstract and full-text screening (91).…”
Section: Methodsmentioning
confidence: 99%
“…Despite progress in the use of machine learning to help automate abstract screening [70,71], we concluded this technology is not sufficiently developed to have been applicable without further testing. Innovative methods are much needed, particularly as the size and scope of maps and reviews continue to increase [72,73].…”
Section: Efficiency In the Review Processmentioning
confidence: 99%
“…Following the completion of all searches, these will be imported into Colandr (a product of a collaborative partnership between the Science for Nature and People Partnership Evidence-Based Conservation working group, DataKind, and Conservation International-https ://www.colan drapp .com/) to allow screening by multiple screeners. Colandr provides a platform for all systematic mapping requirements, from title and abstract screening to meta-data extraction, and applies machine learning and natural language processing algorithms to sort articles according to relevance which can reduce the time taken to screen compared with traditional methods [25]. An exercise to measure inter-rater agreement using Cohen's kappa [26] will be undertaken before title, abstract and full-text screening and meta-data extraction [27].…”
Section: Screening Processmentioning
confidence: 99%