Study on the legal framework of text and data mining (TDM)

Jean - Paul Triaille | March 2014

Publié le 24 avril 2014, par Thérèse Hameau

Excerpts of Conclusion and Summary

The purpose of this Study was to describe how data analysis fits within the present legal context in Europe (both in terms of copyright and of database protection), to highlight the issues which may constitute obstacles or difficulties when applying the existing legal texts, and to analyze whether a new exception for data analysis would be useful or necessary. The purpose was not to suggest ready-to-use provisions for a new possible directive but only to give directions and make suggestions.

We first tried to propose a definition of TDM. Regarding terminology, we suggest referring to “data analysis” rather than to “text and data mining” (TDM). “Data analysis” is also the terminology used in the very few legislative texts or drafts presently existing in or out of the European Union. As a definition, we suggest “ the automated processing of digital materials, which may include texts, data, sounds, images or other elements, or a combination of these, in order to uncover new knowledge or insights”.


On the basis of our analysis, our suggestion is to have a new specific data analysis exception which would be inspired from, and contain partly the same conditions than, the scientific research exceptions, but which would have its own peculiarities. The suggestions are made only for TDM, not for scientific research in general.


What should be the main characteristics and contents of this new exception ?

  • the exception would only apply where the purpose is mainly (and not “solely”) scientific research ;
  • it would not just serve “to illustrate scientific research” but would applymore broadly in cases of “scientific research” ;
  • it would only apply where justified by non-commercial purposes ;
  • mentioning the sources or the names of the authors of the preexisting materials would not be an obligation but be left to the discretion of the researcher(s) ;
  • it would be an exception to the reproduction right (InfoSoc Directive and Database Directive), the adaptation right (both under the Database Direction and as part of the reproduction right in the InfoSoc Directive) and the extraction right (Database Directive) ;
  • it would not apply to tools designed for data analysis (which should remain untouched by the exception) ;
  • it would not apply if the data analysis output substitutes for the pre-existing works or databases and makes the consultation of these pre-existing elements useless ;
  • it would only benefit users having a lawful access to the data ;
  • it could not be overridden by contractual terms ;
  • it would not be optional for Member States but would be mandatory, so as to ensure a level playing field throughout the European Union.

Le rapport