Lex Rosetta: Transfer of Predictive Models Across Languages, Jurisdictions, and Legal Domains

Overview

Lex Rosetta: Transfer of Predictive Models Across Languages, Jurisdictions, and Legal Domains

This is an accompanying repository to the ICAIL 2021 paper entitled "Lex Rosetta: Transfer of Predictive Models Across Languages, Jurisdictions, and Legal Domains". All the data and the code used in the experiments reported in the paper are to be found here.

Data

The data set consists of 807 adjudicatory decisions from 7 different countries (6 languages) annotated in terms of the following type system:

  • Out of Scope - Parts outside of the main document body (e.g., metadata, editorial content, dissents, end notes, appendices).
  • Heading - Typically an incomplete sentence or marker starting a section (e.g., “Discussion,” “Analysis,” “II.”).
  • Background - The part where the court describes procedural history, relevant facts, or the parties’ claims.
  • Analysis - The section containing reasoning of the court, issues, and application of law to the facts of the case.
  • Introductory Summary - A brief summary of the case at the beginning of the decision.
  • Outcome - A few sentences stating how the case was decided (i.e, the overall outcome of the case).

The country specific subsets:

  • Canada - Random selection of cases retrieved from www.canlii.org from multiple provinces. The selection is not limited to any specific topic or court.
  • Czech Republic - A random selection of cases from Constitutional Court (30), Supreme Court (40), and Supreme Administrative Court (30). Temporal distribution was taken into account.
  • France - A selection of cases decided by Cour de cassation between 2011 and 2019. A stratified sampling based on the year of publication of the decision was used to select the cases.
  • Germany - A stratified sample from the federal jurisprudence database spanning all federal courts (civil, criminal, labor, finance, patent, social, constitutional, and administrative).
  • Italy - The top 100 cases of the criminal courts stored between 2015 and 2020 mentioning “stalking” and keyed to the Article 612 bis of the Criminal Code.
  • Poland - A stratified sample from trial-level, appellate, administrative courts, the Supreme Court, and the Constitutional tribunal. The cases mention “democratic country ruled by law.”
  • U.S.A. I - Federal district court decisions in employment law mentioning “motion for summary judgment,” “employee,” and “independent contractor.”
  • U.S.A. II - Administrative decisions from the U.S. Department of Labor. Top 100 ordered in reverse chronological rulings order, starting in October 2020, were selected.

For more detailed information, please, refer to the original paper.

How to Use

ICAIL 2021 Data

The data used in the ICAIL 2021 experiments can be found in the following paths:

data/Country-Language-*/annotator-*-ICAIL2021.csv

Note that the Canadian subset could not be included in this repository due to concerns about personal information protection in Canada. However, it can be obtained upon request at [email protected]. Once you obtain the data, you just need to create data/Canada-EN-1 directory and place all the files there.

If you would like to experiment with different preprocessing techniques the original texts are placed in the following paths:

data/Country-Language-*/texts

You can find the annotations corresponding to these texts here:

data/Country-Language-*/annotator-*.csv

The texts cleaned of the Out of Scope and Heading segments (via dataset_clean.py) are placed in the following paths:

data/Country-Language-*/texts-clean-annotator-*

Note that the processing depends on annotations. Hence, there are several versions of documents at this stage if there were multiple annotators. The annotations corresponding to the cleaned texts are here:

data/Country-Language-*/annotator-*-clean.csv

The dataset_ICAIL2021.py has the processing code that has been applied to the cleaned texts and annotations to generate the ICAIL 2021 dataset (see above). Note, that the code will skip the Czech Republic subset by default. This is because this subset requires an external resource for sentence segmentation (czech-pdt-ud-X.X-XXXXXX.udpipe). You first need to obtain the file at https://universaldependencies.org/. Then, you need to place it into the data directory. Then, you can remove the Czech_Republic-CZ-1 string from the EXCLUDED tuple in dataset_ICAIL2021.py. Finally, you need to replace the data/czech-pdt-ud-2.5-191206.udpipe string in the utils.py to correspond to the file that you have downloaded. After these changes, the code will also operate on the Czech Republic part of the dataset.

Dataset Statistics

To replicate the inter-annotator agreement analysis performed in the ICAIL 2021 paper you can use the ia_agreement.ipynb notebook.

To generate the dataset statistics reported in the ICAIL 2021 paper you can use the dataset_statistics.ipynb notebook.

Experiments

The file ICAIL2021_experiments.ipynb contains the code necessary to run the code presented in the paper. This includes the code to embed the sentences of the cases into a multilingual vector representation, the definition of the Gated Recurrent Unit model and the code to train and evaluated along the different experiments described in the paper. It also contains the code to create the visualizations presented in the discussion section of the paper.

The notebook can be run in two different ways:

Attribution

We kindly ask you to cite the following paper:

@inproceedings{savelka2021,
    title={Lex Rosetta: Transfer of Predictive Models Across Languages, Jurisdictions, and Legal Domains},
    author={Jaromir Savelka and Hannes Westermann and Karim Benyekhlef and Charlotte S. Alexander and Jayla C. Grant and David Restrepo Amariles and Rajaa El Hamdani and S\'{e}bastien Mee\`{u}s and Aurore Troussel and Micha\l\ Araszkiewicz and Kevin D. Ashley and Alexandra Ashley and Karl Branting and Mattia Falduti and Matthias Grabmair and Jakub Hara\v{s}ta and Tereza Novotn\'a, Elizabeth Tippett and Shiwanni Johnson},
    year={2021},
    booktitle={Proceedings of the 18th International Conference on Artificial Intelligence and Law},
    publisher={Association for Computing Machinery},
    doi={10.1145/3462757.3466149}
}

Jaromir Savelka, Hannes Westermann, Karim Benyekhlef, Charlotte S. Alexander, Jayla C. Grant, David Restrepo Amariles, Rajaa El Hamdani, Sébastien Meeùs, Aurore Troussel, Michał Araszkiewicz, Kevin D. Ashley, Alexandra Ashley, Karl Branting, Mattia Falduti, Matthias Grabmair, Jakub Harašta, Tereza Novotná, Elizabeth Tippett, and Shiwanni Johnson. 2021. Lex Rosetta: Transfer of Predictive Models Across Languages, Jurisdictions, and Legal Domains. In Eighteenth International Conference for Artificial Intelligence and Law (ICAIL’21), June 21–25, 2021, São Paulo, Brazil. ACM, New York,NY, USA, 10 pages. https://doi.org/10.1145/3462757.3466149

Official code for CVPR2022 paper: Depth-Aware Generative Adversarial Network for Talking Head Video Generation

📖 Depth-Aware Generative Adversarial Network for Talking Head Video Generation (CVPR 2022) 🔥 If DaGAN is helpful in your photos/projects, please hel

Fa-Ting Hong 503 Jan 04, 2023
[ICCV'21] Official implementation for the paper Social NCE: Contrastive Learning of Socially-aware Motion Representations

CrowdNav with Social-NCE This is an official implementation for the paper Social NCE: Contrastive Learning of Socially-aware Motion Representations by

VITA lab at EPFL 125 Dec 23, 2022
UMT is a unified and flexible framework which can handle different input modality combinations, and output video moment retrieval and/or highlight detection results.

Unified Multi-modal Transformers This repository maintains the official implementation of the paper UMT: Unified Multi-modal Transformers for Joint Vi

Applied Research Center (ARC), Tencent PCG 84 Jan 04, 2023
A foreign language learning aid using a neural network to predict probability of translating foreign words

Langy Langy is a reading-focused foreign language learning aid orientated towards young children. Reading is an activity that every child knows. It is

Shona Lowden 6 Nov 17, 2021
Preprossing-loan-data-with-NumPy - In this project, I have cleaned and pre-processed the loan data that belongs to an affiliate bank based in the United States.

Preprossing-loan-data-with-NumPy In this project, I have cleaned and pre-processed the loan data that belongs to an affiliate bank based in the United

Dhawal Chitnavis 2 Jan 03, 2022
This project provides the code and datasets for 'CapSal: Leveraging Captioning to Boost Semantics for Salient Object Detection', CVPR 2019.

Code-and-Dataset-for-CapSal This project provides the code and datasets for 'CapSal: Leveraging Captioning to Boost Semantics for Salient Object Detec

lu zhang 48 Aug 19, 2022
Display, filter and search log messages in your terminal

Textualog Display, filter and search logging messages in the terminal. This project is powered by rich and textual. Some of the ideas and code in this

Rik Huygen 24 Dec 10, 2022
Deep Structured Instance Graph for Distilling Object Detectors (ICCV 2021)

DSIG Deep Structured Instance Graph for Distilling Object Detectors Authors: Yixin Chen, Pengguang Chen, Shu Liu, Liwei Wang, Jiaya Jia. [pdf] [slide]

DV Lab 31 Nov 17, 2022
RRL: Resnet as representation for Reinforcement Learning

Resnet as representation for Reinforcement Learning (RRL) is a simple yet effective approach for training behaviors directly from visual inputs. We demonstrate that features learned by standard image

Meta Research 21 Dec 07, 2022
Geneva is an artificial intelligence tool that defeats censorship by exploiting bugs in censors

Geneva is an artificial intelligence tool that defeats censorship by exploiting bugs in censors

Kevin Bock 1.5k Jan 06, 2023
NeuroGen: activation optimized image synthesis for discovery neuroscience

NeuroGen: activation optimized image synthesis for discovery neuroscience NeuroGen is a framework for synthesizing images that control brain activatio

3 Aug 17, 2022
机器学习、深度学习、自然语言处理等人工智能基础知识总结。

说明 机器学习、深度学习、自然语言处理基础知识总结。 目前主要参考李航老师的《统计学习方法》一书,也有一些内容例如XGBoost、聚类、深度学习相关内容、NLP相关内容等是书中未提及的。

Peter 445 Dec 12, 2022
Discerning Decision-Making Process of Deep Neural Networks with Hierarchical Voting Transformation

Configurations Change HOME_PATH in CONFIG.py as the current path Data Prepare CENSINCOME Download data Put census-income.data and census-income.test i

2 Aug 14, 2022
Official PyTorch implementation of DD3D: Is Pseudo-Lidar needed for Monocular 3D Object detection? (ICCV 2021), Dennis Park*, Rares Ambrus*, Vitor Guizilini, Jie Li, and Adrien Gaidon.

DD3D: "Is Pseudo-Lidar needed for Monocular 3D Object detection?" Install // Datasets // Experiments // Models // License // Reference Full video Offi

Toyota Research Institute - Machine Learning 364 Dec 27, 2022
[PAMI 2020] Show, Match and Segment: Joint Weakly Supervised Learning of Semantic Matching and Object Co-segmentation

Show, Match and Segment: Joint Weakly Supervised Learning of Semantic Matching and Object Co-segmentation This repository contains the source code for

Yun-Chun Chen 60 Nov 25, 2022
JAX-based neural network library

Haiku: Sonnet for JAX Overview | Why Haiku? | Quickstart | Installation | Examples | User manual | Documentation | Citing Haiku What is Haiku? Haiku i

DeepMind 2.3k Jan 04, 2023
A Moonraker plug-in for real-time compensation of frame thermal expansion

Frame Expansion Compensation A Moonraker plug-in for real-time compensation of frame thermal expansion. Installation Credit to protoloft, from whom I

58 Jan 02, 2023
Real-time Object Detection for Streaming Perception, CVPR 2022

StreamYOLO Real-time Object Detection for Streaming Perception Jinrong Yang, Songtao Liu, Zeming Li, Xiaoping Li, Sun Jian Real-time Object Detection

Jinrong Yang 237 Dec 27, 2022
Code for the paper "Zero-shot Natural Language Video Localization" (ICCV2021, Oral).

Zero-shot Natural Language Video Localization (ZSNLVL) by Pseudo-Supervised Video Localization (PSVL) This repository is for Zero-shot Natural Languag

Computer Vision Lab. @ GIST 37 Dec 27, 2022
Readings for "A Unified View of Relational Deep Learning for Polypharmacy Side Effect, Combination Therapy, and Drug-Drug Interaction Prediction."

Polypharmacy - DDI - Synergy Survey The Survey Paper This repository accompanies our survey paper A Unified View of Relational Deep Learning for Polyp

AstraZeneca 79 Jan 05, 2023