BRIC_2024_2024-06-06-11:30-haunschild_archival_version.pdf

These are the slides of my presentation at BRIC 2024 about global science overlay maps using OpenAlex.

1. Global Science Overlay Maps
Robin Haunschild
BRIC 2024, Vancouver, June 06, 2024

2. Global maps
Global vs. local
Local maps only show the nodes of the focus dataset
Global maps show the nodes of the focus dataset embedded in the
environment
Node positions are different for local and global maps
Local maps are “cheaper”
Global maps are “more expensive”
Global Science Overlay Maps R. Haunschild & L. Bornmann 2

3. Overlay maps and base maps
What are overlay maps and base maps?
Overlay maps use a base map to project data/information onto it
Usually, the node size is scaled with the number of papers
Node colors can be associated with impact, cluster assignment, eld
assignment, or other pieces of information
Overlay maps can be constructed from local or global base maps
Base maps provide the coordinates and cluster assignments of the nodes – all
other data are provided by the overlay data
Global Science Overlay Maps R. Haunschild & L. Bornmann 3

4. Why OpenAlex?
Why is OpenAlex suitable for global maps?
OpenAlex is free
!Everyone can obtain and use the data
OpenAlex has a very large coverage of the literature!More of the scientic
environment is included
Global Science Overlay Maps R. Haunschild & L. Bornmann 4

5. History
A short history of global overlay maps
Boyack (2009) proposed the methodology for the rst time using data from Web of
Science
Klavans and Boyack (2009) proposed a global overlay map of the scientic landscape
Description of two general approaches for creating global maps on the basis of
individual documents (Klavans & Boyack, 2010)
The method was developed further by Rafols, Porter, & Leydesdorff (2010),
Leydesdorff & Rafols (2012), and Leydesdorff et al. (2013) – interactive overlays
based on Web of Science were presented
Boyack, K. W. (2009). Using detailed maps of science to identify potential collaborations. Scientometrics, 79(1), 27-44. DOI: DOI
10.1007/s11192-009-0402-6.
Klavans, R. & Boyack, K. W. (2009). Toward a consensus map of science. Journal of the American Society for Information Science and Technology,
60(3), 455-476. DOI: 10.1002/asi.20991.
Klavans, R. & Boyack, K. W. (2010). Using global mapping to create more accurate document-level maps of research elds. Journal of the American
Society for Information Science and Technology, 62(1), 1-18. DOI: 10.1002/asi.21444.
Leydesdorff, L. & Rafols, I. (2012). Interactive overlays: A new method for generating global journal maps from Web of Science data. Journal of
Informetrics, 6(2), 318-332. DOI: 10.1016/j.joi.2011.11.003.
Leydesdorff, L., Rafols, I., & Chen, C. (2013). Interactive overlays of journals and the measurement of interdisciplinarity on the basis of aggregated
journal-journal citations. Journal of the American Society for Information Science and Technology, 64(12), 2573-2586. DOI: 10.1002/asi.22946.
Rafols, I., Porter, A. L., & Leydesdorff, L. (2010). Science Overlay Maps: A New Tool for Research Policy and Library Management. Journal of the
American Society for Information Science and Technology, 61(9), 1871-1887. DOI: 10.1002/asi.21368.
Global Science Overlay Maps R. Haunschild & L. Bornmann 5

6. History
A short history of global overlay maps
Leydesdorff et al. (2012) created global overlay maps with MeSH terms on the basis of
PubMed
Leydesdorff et al. (2015) presented overlay maps based of Scopus data
Bornmann & Haunschild (2016) transferred the method from bibliometrics to altmetrics
by applying it to Mendeley data
VOSviewer (van Eck & Waltman, 2010) is well suited to create and use overlay maps
Kevin Boyack and Richard Klavans received the Derek de Solla Price Memorial Medal
in 2023 as a team “for their outstanding contribution to the eld of scientometrics”
(Editors, 2023). Kevin Boyack highlighted their work on global overlay maps in his
presentation
Bornmann, L. & Haunschild, R. (2016). Overlay maps based on Mendeley data: The use of altmetrics for readership networks. Journal of the
Association for Information Science and Technology, 67(12), 3064-3072. DOI: 10.1002/asi.23569.
Editors (2023). The team Kevin W. Boyack and Richard Klavans wins the 2023 Derek John de Solla Price Medal. Scientometrics. DOI:
10.1007/s11192-023-04847-7.
Leydesdorff, L., Rotolo, D., & Rafols, I. (2012). Bibliometric perspectives on medical innovation using the medical subject Headings of PubMed.
Journal of the American Society for Information Science and Technology, 63(11), 2239-2253. DOI: 10.1002/asi.22715.
Leydesdorff, L., de Moya-Anegon, F., & Guerrero-Bote, V. P. (2015). Journal Maps, interactive overlays, and the measurement of interdisciplinarity on
the basis of Scopus data (1996-2012). Journal of the American Society for Information Science and Technology, 66(5), 1001-1016. DOI:
10.1002/asi.23243.
van Eck, N. J. & Waltman, L. (2010). Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics, 84(2), 523-538. DOI:
10.1007/s11192-009-0146-3.
Global Science Overlay Maps R. Haunschild & L. Bornmann 6

7. Dataset and method
Dataset
OpenAlex snapshot from August 2023 with 243,053,925 documents in the
“Kompetenznetzwerk Bibliometrie” (KB)
1800-2022: 237,876,541 documents
2008-2022: 134,092,007 documents
2013-2022: 95,438,459 documents
2018-2022: 47,665,990 documents
2022: 8,496,167 documents
Global Science Overlay Maps R. Haunschild & L. Bornmann 7

8. Dataset and method
Method – node selection
Nodes are concepts
Levels 0-2 out of 0-5 are used.
19 concepts on level 0, 284 concepts on level 1,
and 21,460 concepts on level 2
1800-2022: 237,830,057 documents with at least one of 21,758 concepts
2008-2022: 134,054,634 documents with at least one of 21,758 concepts
2013-2022: 95,406,638 documents with at least one of 21,758 concepts
2018-2022: 47,641,330 documents with at least one of 21,756 concepts
2022: 8,478,584 documents with at least one of 21,715 concepts
No document type restrictions for obtaining global maps
Global Science Overlay Maps R. Haunschild & L. Bornmann 8

9. Dataset and method
Method – positioning of the nodes
Direct citation relations are used for the calculation of the node positions
Usage of a ve-year citation (i.e., referencing) window (plus publication year)
as default
One map was created using a thirty-year citation window (plus publication year)
select cd.id cited_concept, cg.id citing_concept, count(*) cit_rels from
z_openalex_rep_20230819_openbib.works_referenced_works r join
z_openalex_rep_20230819_openbib.works_concepts wcd on referenced_work_id=wcd.work_id join
z_openalex_rep_20230819_openbib.concepts cd on cd.id=wcd.concept_id join
z_openalex_rep_20230819_openbib.works wd on wd.id=wcd.work_id join
z_openalex_rep_20230819_openbib.works_concepts wcg on r.work_id=wcg.work_id join
z_openalex_rep_20230819_openbib.concepts cg on cg.id=wcg.concept_id join
z_openalex_rep_20230819_openbib.works wg on wg.id=wcg.work_id where cd.level between 0 and 2 and
cg.level between 0 and 2 and wg.publication_year-wd.publication_year between 0 and 5 and wg.publication_year
between 2008 and 2022 group by cd.id, cg.id order by cd.id, cg.id;
Global Science Overlay Maps R. Haunschild & L. Bornmann 9

10. Dataset and method
Method – Creation of the base map
Import SQL result into VOSviewer
„Create”!„Create a map based on network data”!„VOSviewer network le”
Save map (and where required network) le
Replace concept IDs with display names
Load new map le in VOSviewer and adjust parameters, here:
Cluster resolution: 1.25
Min. cluster size: 500
Label size variation: 0.2 (0.4)
Scale: 0.5 (1.0)
Global Science Overlay Maps R. Haunschild & L. Bornmann 10

11. Dataset and method
Method – overlay of data
Export data (papers per concept) from OpenAlex:
Web interfaceAPISnapshot
Fill in the number of papers in the column “weight<papers>” in the row of the
concept if the focal unit has at least one paper assigned to this concept
Remove the values in the columns “label” and “weight<papers>” in the row of
the concept and set the color to a different value (e.g.,) if the focal
unit has no papers assigned to this concept
Global Science Overlay Maps R. Haunschild & L. Bornmann 11

12. Dataset and method
Method – normalized overlay maps
Export number of papers (N) per concept (c) on levellin the OpenAlex world
(w) for the map or focal unit (u)
Nwl=
P
c
Ncwl
Nul=
P
c
Ncul
pcwl=Ncwl=Nwl
pcul=Ncul=Nul
Acul=pcul=pcwl
Fill inAculin the column “weight<papers>” in the row of the concept if the focal
unit has at least one paper assigned to this concept
Remove the values in the columns “label” and “weight<papers>” in the row of
the concept and set the color to a different value (e.g.,) if the focal
unit has no papers assigned to this concept
Global Science Overlay Maps R. Haunschild & L. Bornmann 12

13. Base maps
Six different base maps
A) 1800-2022 with a 30-year citation window
B) 1800-2022 with a ve-year citation window
C) 2008-2022 with a ve-year citation window
D) 2013-2022 with a ve-year citation window
E) 2018-2022 with a ve-year citation window
F) 2022 with a ve-year citation window
Link to base maps: http://ivs.fkf.mpg.de/global_maps_OpenAlex/
Global Science Overlay Maps R. Haunschild & L. Bornmann 13

14. Comparison of the six base maps
A)
B)C)
D)
E)F)
Global Science Overlay Maps R. Haunschild & L. Bornmann 14

15. A) 1800-2022 with a 30-year citation window
Cluster colors were assigned
according to cluster size:
orange (social sciences
and humanities),
(medicine),
ics, computer sciences, and
theoretical physics),
(physics and engineering),
pink (biology), and
(chemistry and material sci-
ences)
Global Science Overlay Maps R. Haunschild & L. Bornmann 15

16. B) 1800-2022 with a ve-year citation window
Cluster colors were assigned
according to cluster size:
orange (social sciences
and humanities),
(medicine),
and engineering),
(mathematics, computer
sciences, and theoretical
physics),, and
light-blue (chemistry and
material sciences)
Global Science Overlay Maps R. Haunschild & L. Bornmann 16

17. C) 2008-2022 with a ve-year citation window
Cluster colors were assigned
according to cluster size:
orange (social sciences
and humanities),
(medicine),
and engineering),
(mathematics, computer
sciences, and theoretical
physics),, and
light-blue (chemistry and
material sciences)
Global Science Overlay Maps R. Haunschild & L. Bornmann 17

18. D) 2013-2022 with a ve-year citation window
Cluster colors were assigned
according to cluster size:
orange (social sciences
and humanities),
(medicine),
ics, computer sciences, and
theoretical physics),
(physics and engineering),
pink (Biologie), and
(chemistry and material sci-
ences)
Global Science Overlay Maps R. Haunschild & L. Bornmann 18

19. E) 2018-2022 with a ve-year citation window
Cluster colors were assigned
according to cluster size:
orange (social sciences
and humanities),
(medicine),
ics, computer sciences, and
theoretical physics),
(physics and engineering),
pink (biology), and
(chemistry and material sci-
ences)
Global Science Overlay Maps R. Haunschild & L. Bornmann 19

20. F) 2022 with a ve-year citation window
Cluster colors were assigned
according to cluster size:
orange (social sciences
and humanities),
(medicine),
ics, computer sciences, and
theoretical physics),
(biology),
engineering), and
(chemistry and material sci-
ences)
Global Science Overlay Maps R. Haunschild & L. Bornmann 20

21. Examples for overlay maps
Six different overlay maps on the base map 2008-2022
A) Papers that are assigned to the OpenAlex author ID of Robin Haunschild
B) Papers that are assigned to the OpenAlex institute ID of the Max Planck Institute for
Solid State Reserach (MPI-FKF)
C) Papers that are assigned to the OpenAlex institute ID of Victoria University
D) Papers that are assigned to the OpenAlex institute ID of Calgary University
E) Papers that are assigned to the OpenAlex institute ID of McGill University
F) Papers that are assigned to the OpenAlex institute ID of McMaster University
G) Papers that are assigned to the OpenAlex institute ID of University of Toronto
H) Papers that are assigned to the OpenAlex institute ID of University of British
Columbia
Global Science Overlay Maps R. Haunschild & L. Bornmann 21

22. Example: global overlay maps of Robin Haunschild
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 22

23. Example: global overlay maps of MPI-FKF
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 23

24. Example: global overlay maps of Victoria University
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 24

25. Example: global overlay maps of Calgary University
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 25

26. Example: global overlay maps of McGill University
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 26

27. Example: global overlay maps of McMaster University
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 27

28. Example: global overlay maps of University of Toronto
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 28

29. Example: global overlay maps of University of British Columbia
(without normalization) (with normalization)
Global Science Overlay Maps R. Haunschild & L. Bornmann 29

30. Example: normalized global overlay maps of six Canadian
universities
Calgary University Victoria University McMaster UniversityMcGill University University of Toronto University of British Columbia
Global Science Overlay Maps R. Haunschild & L. Bornmann 30

31. Summary
Summary
Base maps for different time periods were provided
Citation windows of ve years seem to be sufcient for base maps
Data for overlay maps can be retrieved via the GUI, API, or from a snapshot of
OpenAlex
Wrongly assigned papers (to concepts as well as to units) produce surprising
results – data correction might help
Improvement of the assignment of the papers to concepts in OpenAlex would
improve the base maps and the overlay maps
Global Science Overlay Maps R. Haunschild & L. Bornmann 31

32. Outlook
Outlook
Construction of global overlay maps on the basis of the
new topics in OpenAlex
journals in OpenAlex
subjects in Scilit
Usage of additional data for the overlay color (e.g., publication year, scientic
impact)
Global Science Overlay Maps R. Haunschild & L. Bornmann 32

33. https://ivs.fkf.mpg.de/slides_bric2024.pdf
arXiv preprint:
https://arxiv.org/abs/2404.02732
Email:
[email protected]
Twitter:
@rhaunschild
Mastodon:
@[email protected] 532316532316
Global Science Overlay Maps R. Haunschild & L. Bornmann 33