Download 3.1 Is Europe Falling Behind in Data Mining

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Institute for Information Law
Is Europe falling behind? Copyright’s impact on
data mining in academic research
C. Handke, L. Guibault, J.J. Vallbé, LIBER, 24.06.2015
Fact:
Data mining involves access to and usage of
(copyright protected) articles and data in bulk
Question:
Does copyright law impact data mining in
academic research?
2
Empirical research


Use of data available on academic research
output
Academic publications as reasonable indicator
of the innovation output of academic
researchers
3
Collection of data

...1/
Source of data:
 From Thomson Reuter’s Web of Science
(WoS)

Entire WoS Core Collection Database
including Science Citation Index Expanded,
Social Science Citation Index and Art &
Humanities Citation Index.

For 40 countries, between 1993 and 2014
4
Collection of data

...2/
Use of key words : ‘data mining’

18,441 DM-related articles between 1993 and
2014.

23,802,650 articles for the entire panel.

For all countries and entire time period covered,
0.7‰ had DM as a topic
5
Absolute number of DM research articles published per
year (42 countries, 1992 to 2014)
6
Classification of countries




Copyright law is not harmonised
Protection is determined at national level
4 categories:

Not allowed

Probably not allowed

Probably allowed

Allowed
Few countries remained unclassified
7
Classification in detail


...1/
‘Not allowed’: countries with a closed list of
exception and limitation, without relevant
exception = Europe, Switzerland, Russia, Latin
America
‘Probably not allowed’: countries with a fair
dealing exception, without relevant case law =
Australia, India, Malaysia, Nigeria, Thailand
8
Classification in detail


...2/
‘Probably allowed’: countries with fair use
defence, without relevant case law = Canada,
China, Israel, Korea, Singapore, Taiwan
‘Allowed’: countries with specific relevant
exception or fair use defence with relevant case
law = Japan, United Kingdom, United States
9
Countries that underwent a change in
status between 1992 and 2014

From ‘probably not allowed’ to ‘probably
allowed’

Canada, China, Israel, Korea, Singapore,
Taiwan,

From ‘not/probably not allowed’ to ‘allowed’

Japan (2010), United Kingdom (2014)
10
Control indicators
1. GDP per capita
2. Country population size
3. The level of rule of law
11
Analysis & Results



Multilevel linear regression model with varying
intercepts by country, also known as a random
effects model
In all specifications, we find significant positive
coefficients for the category ‘probably allowed’
(p<.01).
Overall, there is extensive evidence that DM share
is greater in countries with more permissive DMrelated copyright than in the ‘not allowed’ category
of countries.
12
Conclusion


In most EU/EEA Member States, DM-related
copyright protection is comparatively strong.
Our results suggest that the net effect is a
weaker performance of domestic academic
researchers in this increasingly important type
of research
13
Draft paper available at:

Handke, Christian and Guibault, Lucie and
Vallbé, Joan-Josep, Is Europe Falling Behind in
Data Mining? Copyright's Impact on Data
Mining in Academic Research (June 7, 2015).
Available at SSRN:
http://ssrn.com/abstract=2608513 or
http://dx.doi.org/10.2139/ssrn.2608513
Type the footer here
14
Thank you very much!!
For questions/discussion
Christian Handke : [email protected]
Lucie Guibault : [email protected]
Joan Josep Vallbé : [email protected]
15