| Description | Aim
Augment the existing legislative data with fine-grained semantic information in order to facilitate legal data processing. Build a prototype Knowledge Graph (KG) of selected UK legislation in Akoma Ntoso that represents Hohfeldian legal relations and fine-grained semantic information; the KG would support SPARQL querying and legal inference. Contribute to the work of The National Archives and other government departments.
Objectives
1. Ingestion & Parsing: Gather a working corpus of selected legislation in Akoma Ntoso format.
2. LLM Extraction to a Semantic Model: Prompt an LLM to: identify provisions that express Hohfeldian relations, e.g., powers, duties, rights, privileges, liabilities, immunities, disabilities; within the provisions,
identify structured semantic information according to a domain model, e.g., actor, main action, negations, object, and exceptions
3. Semantic Mapping: Represent the extracted information in a legal ontology (aligned with a KG) which represents Hohfeldian relations and related semantic model, serialising the data in RDF (Turtle/JSON-LD).
4. Querying & Reasoning: Exemplify querying and rule-based inferences (e.g., derive liabilities from breached duties; propagate powers to delegated authorities). Provide SPARQL endpoints and reusable query templates.
5. Validation: Validate the semantic outputs (of queries and inferences) with domain experts and the KG wellformedness with SHACL.
Use Cases.
• Compliance analytics: “What duties bind local authorities, and what exceptions apply?”
• Delegation tracking: “Which sections grant the Secretary of State power to make regulations?”
• Change impact: Detect affected duties/powers when an Act is amended.
• RegTech integration: Map obligations to internal controls and evidence.
Target Audiences. Lawyers, public administrators, legal informatics researchers, compliance teams, RegTech vendors, and civic tech organizations.
Impact. Improves transparency and discoverability of legal obligations/powers; enables precise queries, checks inferences and consistency, and automates compliance checks. The project will accelerate research, reduce
manual review costs, and provide interoperable, auditable legal data services. |
|---|