This repository contains the code to reproduce the computational study from the paper by Pedro Hernández Serrano, Kody Moodley, Gijs van Dijck, and Michel Dumontier

One of the challenges in computational legal research is trying to quantitatively assess “relevance” in a network of court decisions. In Scientometrics, the phenomena of “delayed recognition” for scientific publications has been studied. In this connection, the term "sleeping beauty" (SB) was coined to denote an article that received almost no attention immediately after publication, but suddenly receives a burst of citations many years later, these publications can be identified by calculating their so-called Beauty coefficient (B-coefficient). In this contribution, we apply approaches used for identifying SBs to the context of European case law, more specifically to decisions arising from the Court of Justice of the European Union (CJEU). We compared B-coefficients of CJEU cases with their centrality scores using classical algorithms from network analysis, finding that these measures tend to correlate. This makes the B-coefficient a potential candidate for an additional measure of case relevance or significance. We discuss the implications of this that are interesting for legal scholars, acknowledging that future work is required to calibrate the scale of the time variable in the B-coefficient formula for finer-grained application to case law. The setup of our study provides a template and foundation for a set of methodologies for case law analytics that extends the power of traditional network analysis techniques for answering questions about the behavior of European court systems.

For attribution in academic contexts, please cite this work as:

@inproceedings{serrano2020sleeping,
  title = {Sleeping Beauties in Case Law},
  author = {{Hern{\'a}ndez Serrano}, Pedro and Kody Moodley and {van Dijck}, Gijs and Michel Dumontier},
  booktitle = {International Conference on Legal Knowledge and Information Systems, JURIX},
  month = dec,
  year = {2020},
  address = {Prague, Czech Republic}
}

• Data Archiving: Findable data can be indexed in public repositories, making it Accessible

• Data Reusing: Reusable data must be uniquly referenced, making it Findable.

• Data Access Protocols: Interoperable data can be retrieved with an open protocol, making it Accesible

• Terms of Use (for machines): Accessible data contains machine friendly terms of use, making it Reusable.

• Rich Metadata: Interoperable metadata is semantically described with ontology terms (e.g. Dublin Core), making it Findable.

• Data Descriptions (for machines): Interoperable data is adequately described, machine friendly, making it Reusable

• Terms of Use (for humans): Accessible data contains human friendly terms of use, making it Reusable.

• Data Descriptions (for humans): Interoperable data is adequately described, human friendly, making it Reusable

This repository is licensed under the terms of the LGPL-2.1 license.