Recognising Legal Characteristics of the Judgments of the European Court of Justice: Difficult but Not Impossible
Research output: Chapter in Book/Report/Conference proceeding › Book chapter › Research
Standard
Recognising Legal Characteristics of the Judgments of the European Court of Justice : Difficult but Not Impossible. / Contini, Alessandro; Piccolo, Sebastiano; Zurita, Lucia Lopez; Sadl, Urska.
Legal Knowledge and Information Systems. ed. / Enrico Francesconi; Georg Borges; Christoph Sorge. Vol. 362 IOS Press, 2022. p. 164-169.Research output: Chapter in Book/Report/Conference proceeding › Book chapter › Research
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - CHAP
T1 - Recognising Legal Characteristics of the Judgments of the European Court of Justice
T2 - Difficult but Not Impossible
AU - Contini, Alessandro
AU - Piccolo, Sebastiano
AU - Zurita, Lucia Lopez
AU - Sadl, Urska
PY - 2022
Y1 - 2022
N2 - Computers perform remarkably in formerly difficult tasks. This article reports thepreliminary results on the prediction of two characteristics of judgments of the European Court of Justice, which require the knowledge of concepts and doctrinesof European Union law and judicial decision-making: The legal importance (doctrinal outcome) and leeway to the national courts and legislators (deference). Theanalysis relies on 1704 manually labelled judgments and trains a set of classifiersbased on word embedding, LSTM, and convolutional neural networks. While allclassifiers exceed simple baselines, the overall performance is weak. This suggestsfirst, that the models learn meaningful representations of the judgments. Second,machine learning encounters significant challenges in the legal domain. These arisedoe to the small training data, significant class imbalance, and the characteristics ofthe variables requiring external knowledge.The article also outlines directions for future research.
AB - Computers perform remarkably in formerly difficult tasks. This article reports thepreliminary results on the prediction of two characteristics of judgments of the European Court of Justice, which require the knowledge of concepts and doctrinesof European Union law and judicial decision-making: The legal importance (doctrinal outcome) and leeway to the national courts and legislators (deference). Theanalysis relies on 1704 manually labelled judgments and trains a set of classifiersbased on word embedding, LSTM, and convolutional neural networks. While allclassifiers exceed simple baselines, the overall performance is weak. This suggestsfirst, that the models learn meaningful representations of the judgments. Second,machine learning encounters significant challenges in the legal domain. These arisedoe to the small training data, significant class imbalance, and the characteristics ofthe variables requiring external knowledge.The article also outlines directions for future research.
U2 - 10.3233/FAIA220461
DO - 10.3233/FAIA220461
M3 - Book chapter
SN - 978-1-64368-364-5
VL - 362
SP - 164
EP - 169
BT - Legal Knowledge and Information Systems
A2 - Francesconi, Enrico
A2 - Borges, Georg
A2 - Sorge, Christoph
PB - IOS Press
ER -
ID: 342605915