Project: AI‑Driven Legal Search with Knowledge Graphs for Dutch Jurisprudence
Description
[Please note that this project is primarily aimed at Dutch-speaking students, as it requires working with large amounts of Dutch texts.]
AI-powered search has rapidly become the new standard for search engines. However, popular techniques (e.g. AI agents or retrieval-augmented generation), face fundamental limitations when applied in the legal domain. In particular, standard techniques often rely on a static corpus, chunked into isolated pieces of text. In law, however, judgments and legal texts are interconnected through complex cross-references and precedents. Naive application of techniques therefore results in loss of essential context and deeper logics required for effective legal queries.
Knowledge graphs offer a promising path forward. By converting unstructured legal texts and court decisions into a structured graph that can be traversed dynamically (e.g. by an AI agent), we can develop a search engine that retrieves richer, completer, and more relevant answers to legal queries.
In this project, you will develop a knowledge-graph based search engine to answer legal queries, based on real-world Dutch judgements and legal texts. The project offers the unique opportunity to combine NLP and directly contribute to solutions that innovate the accessibility of the Dutch legal system.
Student requirements
Given the both theoretical and practical potential of this project, we are looking for an ambitious student who:
- Has a deep interest in the theoretical foundations of knowledge graphs and state of the art search techniques.
- Is committed to delivering a high-quality system that meets practical demands of legal professionals.
- Is excited to work together directly with consultants at T&T data consultancy as well as legal domain experts from de Rechtspraak (the Judiciary), to ensure the usability and usefulness of the solution.
As you will have to work extensively with large amounts of Dutch texts, this project is primarily aimed at Dutch-speaking students.
About T&T Data Consultancy
This project will take place at T&T Data Consultancy (based in Utrecht) n collaboration with the Rechtspraak (the Dutch Judiciary, based in The Hague). T&T is a specialized, boutique data consultancy based in Utrecht that that develops AI solutions, particularly in the public and healthcare sector. Together with the Dutch Judiciary, T&T is currently pioneering the newest AI techniques to shape the future of information provision in the legal domain.
Details
- Supervisor
-
Hilde Weerts
- External location
- T&T Data Consultancy
- Interested?
- Get in contact