狠狠撸

狠狠撸Share a Scribd company logo

Data w-steamm

?
?
Genoveva Vargas-Solar
Genoveva Vargas-Solar

This document discusses analyzing gender gaps in academic publications. It proposes building a dataset from the Microsoft Academic Graph to analyze authorships in top computer science journals based on gender, geographic location, and field of study. Preliminary analysis found the authorship was 52.49% male, 11.75% female, with 29.39% and 6.36% of unknown or initials-only gender attribution. The work aims to better understand and address gender imbalance through quantitative analysis and profiling of scientists' careers based on professional and social factors.

1 of 24
1
2 The earliest archaeologist ?nds date counting to the Upper Paleolithic Era (some 50000 years ago). As was the case with other mathematical operations, it was developed out of need – in this case to represent the size of the group, the number of animals in a herd and similar things.
- Alternative distribution of opportunities & responsibilities - Give access to individuals to all activities in society according to their interests, capacities and merits
FEMINIST TALK
Genoveva Vargas-Solar Senior Scientist, French Council of Scientific Research, LIG-LAFMIA genoveva.vargas@imag.fr Data for empowering women in STEAMM AICCSA, Women in STEM Session, Abu Dhabi, 5th November, 2019 http://vargas-solar.com/w-stem/
VOCABULARY ? Add Arts and Medicine to be universal .. ? May be the ?rst change of mind …
CURRENT SITUATION
CURRENT SITUATION
9 TAKING AN X-RAY OF GENDER GAP IN ACADEMIA
10 GENDER GAP INDEX IN PUBLICATIONS Quantifying the gender gap may - identify fields that will not reach parity without intervention - reveal under-appreciated biases - inform benchmarks for gender balance among conference speakers, editors, and hiring committees Productivity and impact measures: the established and reference indexes to evaluate the quality of scientists’ careers - social factors have been not modelled and considered to ponder and adjust productivity measures - it seems that parenthood, and particularly motherhood, marriage and divorce, gender balance in research groups, disease, economy, political situation of the country where people work are factors with high impact on scientists productivity
11 RESEARCH QUESTIONS - Which is the gender gap index in research papers published in high impact journals? - Is there a correlation with the geographic location of the institution people are working at and gender gap index? - Is the gender gap proportionally smaller in European paper authorships? - Are there any differences between men and women productivity according to the geographical area? - Are women or men more likely to publish in certain computing areas than in others? APPROACH Relate this quantitative observation with contextual data like the notoriety of the journals of the publications
12 BUILDING A DATA SET Microsoft Academic Graph (MAG) - Heterogeneous graph containing scienti?c publication records, citation relationships among those publications, authors, institutions, journals, conferences, and ?elds of study - Used to power experiences in Bing, Cortana, Word, and Microsoft Academic - Modi?ed weakly Step 1 Step 2 https://www.microsoft.com/en-us/research/project/microsoft-academic-graph/
13 BUILDING A DATA SET - Journals in Computer Science indexed by the 2016 release of JCR in MongoDB à 668 journals including, the name of the journal, the number of cites, the IF number, the Eigen factor, the year and the area (a journal can be indexed in more than one area) 1 Impact factor of a journal is calculated by dividing the number of current year citations to the source items published in that journal during the previous two years 2 The Eigenfactor score is intended to measure the importance of a journal to the scientific community, by considering the origin of the incoming citations. It reflects how frequently an average researcher would access content from that journal.
14 AUTHORSHIPS DATA SET Microsoft Academic Graph (MAG)
15 AUTHORSHIPS DATA SET Institution Country North America Europe Asia South America AfricaOceania
16 AUTHORSHIPS DATA SET: MIND THE GAP Female 11,75% Male 52,49% Unknown gender 29,39% Unknown gender, Initials 6,36% Authorships Gender Distribution http://Genderize.io ?
17 AUTHORSHIPS DATA SET: MIND THE GAP Unknown continent Information Systems (CS IS) ┃Artificial Intelligence (CS AI)┃ Software Engineering (CS SE) ┃ Interdisciplinary Applications (CS INTER) ┃ Theory & Methods (CS TM) ┃Hardware & Architecture (CS HA) ┃Cybernetics (CS CYB)
18 GENDER GAP INDEX Do not embrace ignorance !
19 GENDER GAP INDEX
20 - The gender gap phenomenon in science is complex and calls for research in many different directions - Effective pipelines to automatically determine authors gender by combining public data regarding their profile available in different professional data sets like Linkedin, Wikipaedia, ORCID, Publons and other platforms - Provide gender gap metrics to papers, journals, and editorial boards. - If such metadata were included in the published papers, it could be easier to measure gender gap LESSONS LEARNED
21 FUTURE WORK PROFILING SCIENTISTS PRODUCTION ACCORDING TO PROFESSIONAL AND SOCIAL ASPECTS
22 FUTURE WORK INFLUENTIAL GROUPS HOW ARE WOMEN AND MEN EMPOWERED?
COMPLETING STEAMM HISTORY Myriam Mirzakhani Fields Medal Frances Arnold Nobel Price Chemistry Catherine Hamlin Obstetrician Margaret Hamilton Software Engineer, Project Apollo Frances E. Allen Turing Award Barbara Liskov Turing Award Shafi Goldwasser Turing Award Jennifer Widom ACM Award Grace Hooper COBOL Esther Duflo Nobel Price, Economy Donna Strickland Nobel Price, Physics Alondra de la Parra Orchestra director Australia
Genoveva Vargas-Solar Senior Scientist, French Council of Scientific Research, LIG-LAFMIA genoveva.vargas@imag.fr http://vargas-solar.com/w-stem/

More Related Content

Data w-steamm

  • 1. 1
  • 2. 2 The earliest archaeologist ?nds date counting to the Upper Paleolithic Era (some 50000 years ago). As was the case with other mathematical operations, it was developed out of need – in this case to represent the size of the group, the number of animals in a herd and similar things.
  • 3. - Alternative distribution of opportunities & responsibilities - Give access to individuals to all activities in society according to their interests, capacities and merits
  • 5. Genoveva Vargas-Solar Senior Scientist, French Council of Scientific Research, LIG-LAFMIA genoveva.vargas@imag.fr Data for empowering women in STEAMM AICCSA, Women in STEM Session, Abu Dhabi, 5th November, 2019 http://vargas-solar.com/w-stem/
  • 6. VOCABULARY ? Add Arts and Medicine to be universal .. ? May be the ?rst change of mind …
  • 9. 9 TAKING AN X-RAY OF GENDER GAP IN ACADEMIA
  • 10. 10 GENDER GAP INDEX IN PUBLICATIONS Quantifying the gender gap may - identify fields that will not reach parity without intervention - reveal under-appreciated biases - inform benchmarks for gender balance among conference speakers, editors, and hiring committees Productivity and impact measures: the established and reference indexes to evaluate the quality of scientists’ careers - social factors have been not modelled and considered to ponder and adjust productivity measures - it seems that parenthood, and particularly motherhood, marriage and divorce, gender balance in research groups, disease, economy, political situation of the country where people work are factors with high impact on scientists productivity
  • 11. 11 RESEARCH QUESTIONS - Which is the gender gap index in research papers published in high impact journals? - Is there a correlation with the geographic location of the institution people are working at and gender gap index? - Is the gender gap proportionally smaller in European paper authorships? - Are there any differences between men and women productivity according to the geographical area? - Are women or men more likely to publish in certain computing areas than in others? APPROACH Relate this quantitative observation with contextual data like the notoriety of the journals of the publications
  • 12. 12 BUILDING A DATA SET Microsoft Academic Graph (MAG) - Heterogeneous graph containing scienti?c publication records, citation relationships among those publications, authors, institutions, journals, conferences, and ?elds of study - Used to power experiences in Bing, Cortana, Word, and Microsoft Academic - Modi?ed weakly Step 1 Step 2 https://www.microsoft.com/en-us/research/project/microsoft-academic-graph/
  • 13. 13 BUILDING A DATA SET - Journals in Computer Science indexed by the 2016 release of JCR in MongoDB à 668 journals including, the name of the journal, the number of cites, the IF number, the Eigen factor, the year and the area (a journal can be indexed in more than one area) 1 Impact factor of a journal is calculated by dividing the number of current year citations to the source items published in that journal during the previous two years 2 The Eigenfactor score is intended to measure the importance of a journal to the scientific community, by considering the origin of the incoming citations. It reflects how frequently an average researcher would access content from that journal.
  • 14. 14 AUTHORSHIPS DATA SET Microsoft Academic Graph (MAG)
  • 15. 15 AUTHORSHIPS DATA SET Institution Country North America Europe Asia South America AfricaOceania
  • 16. 16 AUTHORSHIPS DATA SET: MIND THE GAP Female 11,75% Male 52,49% Unknown gender 29,39% Unknown gender, Initials 6,36% Authorships Gender Distribution http://Genderize.io ?
  • 17. 17 AUTHORSHIPS DATA SET: MIND THE GAP Unknown continent Information Systems (CS IS) ┃Artificial Intelligence (CS AI)┃ Software Engineering (CS SE) ┃ Interdisciplinary Applications (CS INTER) ┃ Theory & Methods (CS TM) ┃Hardware & Architecture (CS HA) ┃Cybernetics (CS CYB)
  • 18. 18 GENDER GAP INDEX Do not embrace ignorance !
  • 20. 20 - The gender gap phenomenon in science is complex and calls for research in many different directions - Effective pipelines to automatically determine authors gender by combining public data regarding their profile available in different professional data sets like Linkedin, Wikipaedia, ORCID, Publons and other platforms - Provide gender gap metrics to papers, journals, and editorial boards. - If such metadata were included in the published papers, it could be easier to measure gender gap LESSONS LEARNED
  • 21. 21 FUTURE WORK PROFILING SCIENTISTS PRODUCTION ACCORDING TO PROFESSIONAL AND SOCIAL ASPECTS
  • 22. 22 FUTURE WORK INFLUENTIAL GROUPS HOW ARE WOMEN AND MEN EMPOWERED?
  • 23. COMPLETING STEAMM HISTORY Myriam Mirzakhani Fields Medal Frances Arnold Nobel Price Chemistry Catherine Hamlin Obstetrician Margaret Hamilton Software Engineer, Project Apollo Frances E. Allen Turing Award Barbara Liskov Turing Award Shafi Goldwasser Turing Award Jennifer Widom ACM Award Grace Hooper COBOL Esther Duflo Nobel Price, Economy Donna Strickland Nobel Price, Physics Alondra de la Parra Orchestra director Australia
  • 24. Genoveva Vargas-Solar Senior Scientist, French Council of Scientific Research, LIG-LAFMIA genoveva.vargas@imag.fr http://vargas-solar.com/w-stem/