Handke, Guibault and Vallbe (2015)
Contents
Source Details
Handke, Guibault and Vallbe (2015) | |
Title: | Is Europe Falling Behind in Data Mining? Copyright's Impact on Data Mining in Academic Research |
Author(s): | Handke, C., Guibault, L., Vallbe, J. |
Year: | 2015 |
Citation: | Handke, C., Guibault, L., & Vallbé, J. J. (2015). Is Europe Falling Behind in Data Mining? Copyright's Impact on Data Mining in Academic Research. Copyright's Impact on Data Mining in Academic Research (May 20, 2015). |
Link(s): | Definitive |
Key Related Studies: | |
Discipline: | |
Linked by: |
About the Data | |
Data Description: | Data was collected from Thomson Reuter’s Web of Science, including the entire WoS Core Collection Database, the Science Citation Index Expanded, Social Science Citation Index and Art & Humanities Citation Index. To identify the research output of interest, data was extracted on the number of all published research articles on DM from 42 large economies. The panel includes the 15 largest EU Member States, as well as the 27 largest other economies based on national GDP in 2013 according to the World Bank. The data covers the years 1992 to 2014. WoS includes articles published since 1975. It contains no articles on DM published before 1992. There are 966 country-year observations.
The dependent variable is data mining research output. The main independent variable is copyright. Besides the total research output of countries, the authors used several control variables: (1) GDP per capita as reported by the World Bank World Development Indicators with complete data for the 1992-2013 period; (2) country population size, also from official World Bank data and complete for the entire time period studied; and (3) the level of rule of law as reported by the Worldwide Governance Indicators Project. |
Data Type: | Primary data |
Secondary Data Sources: | |
Data Collection Methods: | |
Data Analysis Methods: | |
Industry(ies): | |
Country(ies): | |
Cross Country Study?: | Yes |
Comparative Study?: | Yes |
Literature review?: | No |
Government or policy study?: | No |
Time Period(s) of Collection: |
|
Funder(s): |
Abstract
This empirical paper discusses how copyright affects data mining (DM) by academic researchers. Based on bibliometric data, we show that where DM for academic research requires the express consent of rights holders: (1) DM makes up a significantly lower share of total research output; and (2) stronger rule-of-law is associated with less DM research. To our knowledge, this is the first time that an empirical study bears out a significant negative association between copyright protection and innovation.
Main Results of the Study
Main results of the study: *Countries in which academic researchers must acquire the express consent of rights holders to conduct lawful Data Mining (DM), exhibit a significantly lower share of DM research output relative to total research output. *The number of research articles is a reasonable indicator of innovation by academic researchers. This may be the first instance where an empirical study identifies a significant negative association between copyright protection and the supply of new copyright works of any type. *Regarding DM research, copyright seems to have a negative net effect on innovation.*Attribution rights are relatively important for academic researchers whereas commercial rights regarding the reproduction, making available and application of research results are less important. Our results on the relationship between DM in academic research and relevant copyright policy may not generalize to other copyright industries.* Incentives to publish data suitable for follow-up research requires further attention
Policy Implications as Stated By Author
Policy implications: *In the case of academic research and DM, the adverse consequences of copyright protection on the creation of new information goods are greater than the benefits. *DM research often draws on many input works to which others hold copyrights. Copyright exemptions or limitations could promote this type of research, at least to enable DM of input works that have been publicly financed.
Coverage of Study
Datasets
Sample size: | 966 |
Level of aggregation: | publications per country per year |
Period of material under study: | 1992-2014 |