SINFONICA project aims to develop effective and innovative strategies, methods, and tools to engage users, providers, and stakeholders within the Cooperative, Connected, and Automated Mobility (CCAM) ecosystem. This includes citizens, particularly vulnerable users, transport operators, public administrations, service providers, researchers, and vehicle and technology suppliers. The objective is to systematically gather, understand, and organize their needs, desires, and concerns regarding CCAM in a way that is both manageable and actionable. SINFONICA collaboratively creates a decision support tool, named as Knowledge Map Explorer (KME), specifically designed for designers and policymakers to facilitate the seamless and sustainable deployment of CCAM, ensuring inclusivity and equity for all citizens. The proposed tool utilizes an ontological structure that captures interactions and dependencies within the SINFONICA domain. The ontological framework is populated with data from interviews, focus groups, and workshops conducted in four different European contexts.
Through structured data collection from diverse groups — including under-researched demographics such as women, disabled individuals, and rural residents — the Knowledge Map Explorer integrates domain-specific insights into CCAM solutions, fostering accessibility and inclusivity. By utilizing recommendation engines approaches, this tool aims to go one step forward than traditional knowledge mapping, which often captures only explicit, quantitative data. This tool also incorporates qualitative knowledge, thus enriching CCAM solution frameworks with the multifaceted needs and expectations of diverse user profiles.
The architecture, illustrated bellow, supports end users, including policymakers, developers, and researchers, in deploying CCAM solutions that are inclusive, user-oriented, and aligned with CCAM users’ expectations. By mapping and synthesizing knowledge assets, the Knowledge Map Explorer sets a new paradigm for creating, sharing, and implementing domain specific guidance in CCAM.
The system architecture integrates components that enable user interaction with semantic data. At its core is the User Interface Application, where users input queries and view results. Queries are processed by the Query Module and sent to the Fuseki Server, which uses Apache Jena Fuseki to execute SPARQL queries on RDF datasets.
The Recommendation Engine generates suggestions by interacting with the Rule Engine (SWRL) and the Ontology Service (Pellet, Apache Jena). The Rule Engine applies predefined rules based on user preferences and item attributes, while the Ontology Service manages relationships between concepts for reasoning. For instance, it can recommend guidelines to make CCAM systems inclusive for specific demographics. The engine accesses metadata and ontology data stored in RDF format for reasoning and querying.
Protégé supports ontology development and management with tools for editing, API interaction, and ontology storage. Apache Jena complements this with APIs for SPARQL queries, RDF manipulation, and ontology reasoning, using in-memory or TDB storage for large datasets.
Workflow Summary:
- User Interaction:
The user interacts with a UI to select specific categories. The selections are sent to the backend server as a structured request in HTTP query parameters.
- Backend Server Receives Request:
The server (in our case Spring Boot application) parses the user’s request to extract the selected categories
- Query Construction:
Based on the user’s selections, the backend dynamically generates a SPARQL query tailored to the user’s request.
- SPARQL Query Execution:
The backend uses the Apache Jena API to execute the generated SPARQL query against the Fuseki server.
- Pellet Reasoning (Precomputed):
Precomputed Inference: The dataset in Fuseki already includes data inferred using Pellet and SWRL rules, so the reasoning process has been applied before querying.
- Response Parsing:
The query results, returned in SPARQL JSON format, are parsed by the backend using Jena APIs
- Response to UI:
The transformed results are sent back to the UI in a JSON response via HTTP
The following figures demonstrate the graphical user interface of the system:
Authors: Evangelos Tsougiannis, Giannis Panagiotopoulos, Konstantinos Fokeas, Maria Krikochoriti (ICCS)
