Published September 28, 2023 | Version v1
Software documentation Open

Testing the Bergen-approach search strings for SDG research mapping (v. 1.2.2)

  • 1. University of Bergen
  • 2. University of Stavanger
  • 3. Natural Resources Institute Finland (LUKE)
  • 4. Westen Norway University of Applied Sciences

Description

Here we document a testing process we have carried out for the Bergen-approach search strings for mapping published scientific research related to 10 of the UNs Sustainable Development Goals (Armitage C S, Bjerkan HM, Byholm LP, Gåsemyr I, Lorenz M, & Seland EH. (2023). Search strings for finding SDG related research, Bergen approach. Zenodo. https://doi.org/10.5281/zenodo.7241689). The aim of this testing was to identify issues that should be fixed in a revision of the search strings.

Precision and recall testing were carried out at the SDG target-level for the search strings (topic approach v1.2.2, Web of Science Core Collection syntax). Precision levels varied from 69-88 % for each SDG, with most around 80 %. Recall levels varied more, with some SDG strings performing well already, while others needed more changes (particularly SDGs 3, 4, 7 and 11). Publications identified as “not relevant” or not found during recall testing were examined for why they were found/not found by the strings. Issues were tracked , evaluated and addressed during a subsequent revision of the strings (August/September 2023).

This upload consists of 5 files. The contents of the testing result files are described in the documentation pdf:

  • Documentation and results of testing (1 pdf file)
  • Precision testing results (1 data table with two alternative file formats: xlsx and csv)
  • Recall testing results (1 data table with two alternative file formats: xlsx and csv)

Files

Documentation_Testing_the_Bergen_approach_search_strings_for_SDG_research_mapping_v1-2-2.pdf

Files (718.3 kB)

Name Size Download all
md5:f16ffe31ce9d6519b789b3d707a33af5
322.7 kB Preview Download
md5:cead9629e25042344f0d11537839e4d2
87.0 kB Preview Download
md5:375e840a4c2970db7fede486837c3bfc
65.2 kB Download
md5:9b750a7dbaa12f5af0d7b5dd093c1e3d
157.8 kB Preview Download
md5:03f921bf47f7e7687fa4fe4b4d123be4
85.7 kB Download

Additional details

Related works

Is supplement to
10.5281/zenodo.7241689 (DOI)