Brněnské Pyvo – Právní

The presentation will focus on Legal Judgment Prediction using Neural Language Processing and Machine Learning using PyTorch. The focus will be on leveraging the state of art embedding models to handle tasks such as document classification and prediction based on publicly available district court rulings.

A variety of models will be explored, including OpenAI's text-embedding-3-small and text-embedding-3-large, as well as open-source options like avsolatorio/GIST-small-Embedding-v0 and sentence-transformers/all-MiniLM-L12-v2. The session will also explore the performance of embeddings within Retrieval-Augmented Generation (RAG) over legal text datasets, emphasising how these models manage context search and classification in legal corpora and what are their limitations.

About Pavol:

Pavol Travnik holds a Master's degree in Law and is currently pursuing a Master's in Software Engineering and Big Data. With a multidisciplinary background, Pavol has gained experience in technology, including machine learning and legal frameworks, and has worked on software engineering and quality assurance in multiple roles. His expertise bridges legal theory and computational approaches, driving innovation in legal research.