Overview / Usage

I'm researching natural language processing techniques applied to the handling of court cases. I am using technologies such as BERT, Word2Vec, Clustering, Classification. In NLP I am working on techniques like Argumentation and Semantics. 

Methodology / Approach

We are developing different text representation mechanisms in order to find semantic cohesion between elements of legal texts. We are experimenting with graph representation techniques and word embeddings. In this way, we generate different models of text representation. With such representations we use emsables of classifiers and / or clusterings to test our representations.

Technologies Used

Tensorflow

BERT

Word2Vec

Sckit-learn

Repository

https://github.com/dimmykarson/argGCN