Improving Law Interpretability Using NLP

Explore how Natural Language Processing techniques can enhance law interpretability in this 37-minute Strange Loop Conference talk. Dive into a collaborative project between Bardess Data Scientists and the Government of Ontario, focusing on the Accessibility for Ontarians with Disabilities Act (AODA). Learn about a multi-stage analysis approach that combines NLP methodologies to automate rule extraction, identify compliance entities, and organize legal information. Discover a framework for simplifying legal text accessibility and highlighting areas needing clearer language. Follow along as speaker Serena Peruzzo, a senior data scientist at Bardess Group, guides you through the project background, challenges, and technical aspects, including grammar refreshers, NLP pipeline, and specific use cases. Gain insights into innovative applications of data science in the legal domain and their potential impact on public understanding of laws.

Introduction
Project background
Definition of law
Challenges
Framework
Grammar Refresher
Technology
NLP Pipeline
Example
Parts of Speech
Use Case
Three stages
Step 1 Identify burdens
Performance metric
Subject
Normalization
Vector Representation
Spectrum Embedding
KMeans
TFIDF
Organization
Questions