Text and Data Mining in Intellectual Property Law: Towards an Autonomous Classification of Computational Legal Methods

Text and Data Mining (TDM) can generally be defined as the “process of deriving high-quality information from text and data” and commonly refers to a set of automated analytical tools able to extract new, often hidden, knowledge from existing information. The impact that TDM may have on science, arts and humanities is invaluable. This is because by identifying the correlations and patterns that are often concealed to the eye of a human observer due to the amount, complexity, or variety of data surveyed, TDM allows for the discovery of concepts or the formulation of correlations that would have otherwise remained concealed or undiscovered. Considering this point of view, it can be effectively argued that TDM creates new knowledge from old data. The first part of this paper illustrates the state of the art in the still nascent field of TDM and related technologies applied to IP and legal research more generally. Furthermore, it formulates some proposals of systematic classification in a field that suffers from a degree of terminological vagueness. In particular, this paper argues that TDM, together with other types of data-driven analytical tools, deserves its own autonomous methodological classification as ‘computational legal methods.’ The second part of the chapter paper offers concrete examples of the application of computational approaches in IP legal research. This is achieved by discussing a recent project on TDM, which required the development of different methods in order to address certain problems that emerged during the implementation phase. The discussion allows to take a detailed view of the technology involved, of the relevant skills that legal researchers necessitate, and of the obstacles that the application of TDM to IP research contributes to overcome. In particular, it demonstrates some of the advantages in terms of automation and predictive analysis that TDM enables. At the same time, the limited success of the experiment also shows that there are a number of training and skill-related issues that legal researchers and practitioners interested in the application of TDM and computational legal methods in the field of IP law should consider. Accordingly, the second argument advanced in this chapter paper is that law school programmes should include in their educational offer training on computational legal methods