Copyright's impact on data mining in academic research

Abstract

With the proliferation of digital data, data mining (DM)¿in the sense of the discoveryof valuable structures in large sets of data¿is expected to increase the productivityof many types of research. This paper discusses how copyright affects DM by aca-demic researchers. In some territories, academic DM is lawful if researchers havelawful access to input works. In other territories such as the European Union, lawfulDM additionally requires specific consent by rights holders. Based on bibliometricdata and quasi-experimental research designs, we show that where academic DMrequires specific rights holder consent: (1) DM publications make up a significantlylower share of total research output, and (2) stronger rule of law is associated withless DM research. To our knowledge, this study is the first to empirically documentan adverse effect of intellectual property (IP) on innovation under particular circum-stances. There is strong evidence that copyright exceptions or limitations promotethe adoption of DM research.