← Back to Research
Ontology for digital inclusion and service access in rural environments
Elysium λ Development & Research
I.
Research Question
RQ
How can an ontology-based approach and a recommender (rules + ML) improve the localization and prioritization of digital inclusion resources/essential services in a rural context, minimizing ethical risks (bias, exclusion, opacity) and maximizing social utility?
II.
General Objective
Design and evaluate a reproducible prototype (knowledge graph + recommendation) for a local case, with an explicit framework for ethical evaluation and social utility.
III.
Specific Objectives
- Define a conceptual model (minimal ontology) for "resources/services + user profile + access barriers".
- Build a small knowledge graph (RDF or equivalent structure) with verifiable public data.
- Implement a simple recommender (e.g., transparent rules + ranking) and justify why this design is "appropriate technology" (not oversized).
- Evaluate: coverage, semantic coherence, basic explainability and risk scenarios (who is left out, why).
- Write a mini‑paper with a review of the state of the art and ethical discussion.
IV.
Methodology (Proposal)
- Literature review (ontologies for services; responsible recommendation; digital inclusion; "AI for social good").
- Ontology design (competency questions + iteration).
- Implementation of the KG + query endpoint (even if local).
- Technical evaluation: consistency tests + coverage metrics + query cases.
- Socio‑technical evaluation: risk matrix (bias, accessibility, privacy) using UNESCO/OECD frameworks and the "AI and rights" approach.
V.
Data Sources
- Public lists of services, associations, municipal resources, training, libraries, etc. (light web scraping or careful manual extraction).
- OpenStreetMap as a geographical base and POIs (if maps are included).
- A small corpus of synthetic "needs" to test user profiles.