A geographical analysis of knowledge production in computer science l.jpg
1 / 55

A Geographical Analysis of Knowledge Production in Computer Science.


82 views
Uploaded on:
Category: General / Misc
Description
A Geographical Analysis of Knowledge Production in Computer Science Guilherme Vale Menezes Nivio Ziviani Alberto H. F. Laender Virgílio Almeida gmenezes@dcc.ufmg.br nivio@dcc.ufmg.br laender@dcc.ufmg.br virgilio@dcc.ufmg.br Federal University of Minas Gerais - Brazil Summary
Transcripts
Slide 1

A Geographical Analysis of Knowledge Production in Computer Science Guilherme Vale Menezes Nivio Ziviani Alberto H. F. Laender Virgã­lio Almeida gmenezes@dcc.ufmg.br nivio@dcc.ufmg.br laender@dcc.ufmg.br virgilio@dcc.ufmg.br Federal University of Minas Gerais - Brazil

Slide 2

Summary Introduction Data Gathering Results Conclusions LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 3

The Problem Study the attributes of analysts of Computer Science graduate projects 30 graduate projects in 3 geographic locales Build joint effort interpersonal organizations in view of DBLP We utilize a few measurements of cooperation informal organizations Giant Component Clustering Coefficient LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 4

Steps Comparison between 30 programs in 3 districts Comparison between 30 Computer Science fields Study of the interrelationship between fields Temporal examination of the 3 locales and the fields LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 5

Collaboration Network Author Collaboration LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 6

Collaborations in DCC-UFMG LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 7

Collaborations in DCC-UFMG LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 8

Summary Introduction Data Gathering Results Conclusions LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 9

Data Gathering Part of our information originated from Perfil-CC venture Objective of Perfil-CC : study Brazilian Computer Science graduate projects An arrangement of 30 projects was picked Focus : correlation with N orth American projects Results bolstered open strategies Data accumulated in June 2007 LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 10

Brazilian Programs PUC-Rio, UFRJ, UFPE, UFMG, USP-SP, USP-SC, UNICAMP, UFRGS 8 graduate projects 391 creators LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 11

Canadian and US Programs British Columbia, Toronto, Waterloo, Brown, CalTech, CMU, Cornell, Harvard, Illinois, MIT, Princeton, Stanford, UC Berkeley, UTexas Austin, Washington, Wisconsin 16 graduate projects 1,262 creators LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 12

French , Swiss and UK Programs ETH Zurich, Cambridge U., Imperial College, Oxford U., École Polytechnique, Paris VI 6 graduate projects 611 creators LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 13

Data Gathering Professors got from the departments’ sites Publications from DBLP Programs: 30 Professors: 2,007 Authors: 76,537 Papers: 352,766 Venues: 2,176 LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 14

Venue Classification 2,176 were ordered ( by people ) into 30 handle The rundown of fields was gotten from a survey The brazilian Computer Science research group was counseled 312 scientists recognized 30 unique fields LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 15

Computer Science Fields LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 16

Computer Science Fields Algorithms and Theory LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 17

Computer Science Fields Information Retrieval LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 18

Computer Science Fields Bioinformatics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 19

Summary Introduction Data Gathering Results Conclusions LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 20

General Statistics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 21

General Statistics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 22

General Statistics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 23

General Statistics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 24

Giant Component An associated segment is a most extreme associated subgraph LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 25

Giant Component An associated segment is a greatest associated subgraph LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 26

Giant Component A joined segment is a most extreme joined subgraph The biggest joined segment is the titan segment Giant Component size = 5/11 = 0.45 = 45% LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 27

Giant Component LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 28

Giant Component inside Programs LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 29

Clustering Coefficient LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 30

Clustering Coefficient Clustering coefficient of the system is the normal bunching coefficient of its vertexes The bunching coefficient is a measure of transitivity LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 31

Clustering Coefficient LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 32

Computer Science Fields Clustering Coefficient beneath the normal (87%) for fields firmly identified with Mathematics Algorithms and Theory (79%) Operational Reaseach and Optimization (83%) Formalisms, Logics and Semantics (83%) LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 33

Interrelationship between Fields LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 34

Interrelationship between Fields LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 35

Giant Component Evolution LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 36

Giant Component Evolution LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 37

Giant Component Evolution Increase in the quantity of graduate projects in 1990s LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 38

Giant Component Evolution Increase in government subsidizing LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 39

Giant Component Evolution A movement in strategy: more backing to research bunches rather than people LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 40

Giant Component Evolution LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 41

Giant Component Evolution LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 42

Giant Component Evolution LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 43

Edges versus Vertices LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 44

Clustering Coefficient Evolution 2 entrenched fields Computer Architecture Databases 2 rising fields Bioinformatics Geoinformatics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 45

Clustering Coefficient Evolution LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 46

Clustering Coefficient Evolution Densification LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 47

Summary Introduction Data Gathering Results Conclusions LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 48

Conclusions Analysis of the qualities of specialists of Computer Science graduate projects LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 49

Conclusions Analysis of the attributes of specialists of Computer Science graduate projects Differences in the coordinated effort system of Br, Ca-US and Fr-Sw-UK Giant segment Clustering coefficient LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 50

Conclusions Analysis of the qualities of specialists of Computer Science graduate projects Differences in the cooperation system of Br, Ca-US and Fr-Sw-UK Giant segment Clustering coefficient Smaller bunching coefficient for zones all the more firmly identified with Mathematics LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 51

Conclusions Fast development of the goliath segment in Brazil LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 52

Conclusions Fast development of the monster segment in Brazil The quantity of edges becomes speedier than the quantity of vertices in the three areas; quicker development in Ca-US LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 53

Conclusions Fast development of the titan segment in Brazil The quantity of edges becomes speedier than the quantity of vertices in the three districts; quicker development in Ca-US Densification of rising fields LAboratory for Treating INformation (LATIN) – UFMG - Brazil

Slide 54

References Laender, Lucena, Maldonado, Souza e Silva, Ziviani. Evaluating the Research and Education Quality of the Top Brazilian Graduate Programs. ACM SIGCSE Bulletin , 40:135-145, June 2008. Martins, Gonã§alves, Laender, Ziviani. Surveying the Quality of Scientific Conferences Based on Bibliographic Citations. Scientometrics , to show up. 2009. Research center for Treating INformation (LATIN) – UFMG - Brazil

Slide 55

? Guilherme Vale Menezes Nivio Ziviani Alberto H. F. Laender V