Recognizing cited facts and principles in legal judgements

Olga Shulayeva, Advaith Siddharthan, Adam Wyner (Corresponding Author)

Research output: Contribution to journalArticle

13 Citations (Scopus)
3 Downloads (Pure)

Abstract

In common law jurisdictions, legal professionals cite facts and legal prin- ciples from precedent cases to support their arguments before the court for their in- tended outcome in a current case. This practice stems from the doctrine of stare decisis, where cases that have similar facts should receive similar decisions with respect to the principles. It is essential for legal professionals to identify such facts and principles in precedent cases, though this is a highly time intensive task. In this paper, we present studies that demonstrate that human annotators can achieve reasonable agreement on which sentences in legal judgements contain cited facts and principles (respectively, κ = 0.65 and κ = 0.95 for inter- and intra-annotator agreement). We further demonstrate that it is feasible to automatically annotate sentences containing such legal facts and principles in a supervised machine learning framework based on linguistic features, reporting per category precision and recall figures of between 0.79 and 0.89 for classifying sentences in legal judgements as cited facts, principles or neither using a Bayesian classifier, with an overall κ of 0.72 with the human-annotated gold standard.
Original languageEnglish
Pages (from-to)107-126
Number of pages20
JournalArtificial Intelligence and Law
Volume25
Issue number1
Early online date11 Mar 2017
DOIs
Publication statusPublished - Mar 2017

Fingerprint

gold standard
common law
Linguistics
doctrine
Learning systems
jurisdiction
Classifiers
linguistics
learning
time

Keywords

  • natural language processing
  • legal text analysis
  • legal judgements
  • citations

ASJC Scopus subject areas

  • Artificial Intelligence
  • Law

Cite this

Recognizing cited facts and principles in legal judgements. / Shulayeva, Olga; Siddharthan, Advaith; Wyner, Adam (Corresponding Author).

In: Artificial Intelligence and Law, Vol. 25, No. 1, 03.2017, p. 107-126.

Research output: Contribution to journalArticle

Shulayeva, Olga ; Siddharthan, Advaith ; Wyner, Adam. / Recognizing cited facts and principles in legal judgements. In: Artificial Intelligence and Law. 2017 ; Vol. 25, No. 1. pp. 107-126.
@article{adbbfafab21c447eb035a8ac9d30367e,
title = "Recognizing cited facts and principles in legal judgements",
abstract = "In common law jurisdictions, legal professionals cite facts and legal prin- ciples from precedent cases to support their arguments before the court for their in- tended outcome in a current case. This practice stems from the doctrine of stare decisis, where cases that have similar facts should receive similar decisions with respect to the principles. It is essential for legal professionals to identify such facts and principles in precedent cases, though this is a highly time intensive task. In this paper, we present studies that demonstrate that human annotators can achieve reasonable agreement on which sentences in legal judgements contain cited facts and principles (respectively, κ = 0.65 and κ = 0.95 for inter- and intra-annotator agreement). We further demonstrate that it is feasible to automatically annotate sentences containing such legal facts and principles in a supervised machine learning framework based on linguistic features, reporting per category precision and recall figures of between 0.79 and 0.89 for classifying sentences in legal judgements as cited facts, principles or neither using a Bayesian classifier, with an overall κ of 0.72 with the human-annotated gold standard.",
keywords = "natural language processing, legal text analysis, legal judgements, citations",
author = "Olga Shulayeva and Advaith Siddharthan and Adam Wyner",
note = "Open access via Springer Compact Agreement",
year = "2017",
month = "3",
doi = "10.1007/s10506-017-9197-6",
language = "English",
volume = "25",
pages = "107--126",
journal = "Artificial Intelligence and Law",
issn = "0924-8463",
publisher = "Springer Netherlands",
number = "1",

}

TY - JOUR

T1 - Recognizing cited facts and principles in legal judgements

AU - Shulayeva, Olga

AU - Siddharthan, Advaith

AU - Wyner, Adam

N1 - Open access via Springer Compact Agreement

PY - 2017/3

Y1 - 2017/3

N2 - In common law jurisdictions, legal professionals cite facts and legal prin- ciples from precedent cases to support their arguments before the court for their in- tended outcome in a current case. This practice stems from the doctrine of stare decisis, where cases that have similar facts should receive similar decisions with respect to the principles. It is essential for legal professionals to identify such facts and principles in precedent cases, though this is a highly time intensive task. In this paper, we present studies that demonstrate that human annotators can achieve reasonable agreement on which sentences in legal judgements contain cited facts and principles (respectively, κ = 0.65 and κ = 0.95 for inter- and intra-annotator agreement). We further demonstrate that it is feasible to automatically annotate sentences containing such legal facts and principles in a supervised machine learning framework based on linguistic features, reporting per category precision and recall figures of between 0.79 and 0.89 for classifying sentences in legal judgements as cited facts, principles or neither using a Bayesian classifier, with an overall κ of 0.72 with the human-annotated gold standard.

AB - In common law jurisdictions, legal professionals cite facts and legal prin- ciples from precedent cases to support their arguments before the court for their in- tended outcome in a current case. This practice stems from the doctrine of stare decisis, where cases that have similar facts should receive similar decisions with respect to the principles. It is essential for legal professionals to identify such facts and principles in precedent cases, though this is a highly time intensive task. In this paper, we present studies that demonstrate that human annotators can achieve reasonable agreement on which sentences in legal judgements contain cited facts and principles (respectively, κ = 0.65 and κ = 0.95 for inter- and intra-annotator agreement). We further demonstrate that it is feasible to automatically annotate sentences containing such legal facts and principles in a supervised machine learning framework based on linguistic features, reporting per category precision and recall figures of between 0.79 and 0.89 for classifying sentences in legal judgements as cited facts, principles or neither using a Bayesian classifier, with an overall κ of 0.72 with the human-annotated gold standard.

KW - natural language processing

KW - legal text analysis

KW - legal judgements

KW - citations

U2 - 10.1007/s10506-017-9197-6

DO - 10.1007/s10506-017-9197-6

M3 - Article

VL - 25

SP - 107

EP - 126

JO - Artificial Intelligence and Law

JF - Artificial Intelligence and Law

SN - 0924-8463

IS - 1

ER -