Published June 24, 2015 | Version v1

3.1 Is Europe Falling Behind in Data Mining? Copyright Law's Impact on Data Mining in Academic Research

Description

This paper discusses how different levels of copyright protection affect the text and data mining (TDM) performance of academic researchers in the main research areas.

Copyright protection is determined at a national level. The scope of rights and exceptions varies per country: in some countries, exceptions expressly allow TDM to take place, while in others such activities are restricted. In most countries, the law is unclear. Statutory copyright exceptions, where they exist, can be interpreted in different ways. The assessment of the lawfulness of TDM falls back on the judgment of the researcher. Depending on the knowledge or perception of the law, TDM may be deemed allowed, probably allowed, probably not allowed or restricted. This paper assesses the consequences of the different levels of copyright protection on TDM activities.

Our aim was to explain the comparative variation in research output about data mining. For this, we collected data from Thomson Reuter’s Web of Science. To identify the research output of interest, we extracted all the published research from authors residing in the 31 largest national economies that contained the expression ‘data mining’ in the extended abstract, including 14 EU member states, for the years 1992 to 2014. As a control for the total research output of the respective countries, our dependent variable was the quotient between this absolute academic TDM output and the total research output from these countries. Our unit of analysis was the country-year proportion of TDM research output.

Other control variables included the rule of law (as reported by the World Bank), dealing with the level of enforcement of copyright, and the size and wealth of countries.

To estimate the effect of copyright law on the share of TDM in total research output, we fitted a multilevel linear regression model with varying intercept for country and year.

The data illustrate the rapid growth of TDM-related articles in total research output across all countries.We find a highly significant effect of copyright law: the more restrictive copyright law in most European countries is associated with a significantly lower share of TDM output. Data mining makes up a higher share of total research output in countries with more permissive copyright laws. Some Asian countries in particular over-perform in terms of their TDM research output. What is more, the share of TDM in total research output grows more rapidly in the less restrictive countries.

Lucie Guibault is Associate Professor at the Institute for Information Law in the University of Amsterdam (UvA). She studied law at the Université de Montréal (Canada) and in 2002 she receivedher doctorate from the University of Amsterdam, where she defended her thesis on ‘Copyright Limitations and Contracts: An Analysis of the Contractual Overridability of Limitations on Copyright’. She specialises in international and comparative copyright and intellectual property law, and has carried out research for the European Commission, Dutch ministries, UNESCO and the Council of Europe. Her main areas of interest include copyright and related rights in the information society, open content licensing, collective rights management, limitations and exceptions in copyright, and author’s contract law, She was coauthor of the Report of the Expert Group on Standardisation in the Area of Innovation and Technological Development, Notably in the Field of Text and Data Mining, written for the Directorate- General Research and Innovation, European Commission.

Files

Files (906.6 kB)

Name Size Download all
md5:e2551f29c6a3a706aaa4c80ce8d1de79
906.6 kB Download