Development of knowledge graph for representing industrial processes
Αναπαράσταση βιομηχανικών διεργασιών με χρήση γράφων γνώσης
Keywords
Knowledge graph ; Knowledge representation ; Competency questions ; Wastewater treatment process ; Neo4j ; Cypher ; Γράφημα γνώσης ; Αναπαράσταση γνώσης ; Ερωτήματα επάρκειας ; Διεργασία επεξεργασίας λυμάτωνAbstract
This thesis investigates the design and implementation of knowledge graphs (KGs) using the Neo4j platform, with a focus on modeling industrial wastewater treatment processes. The study emphasizes the advantages of graph-based representations over traditional relational databases, highlighting their flexibility, semantic enrichment, and ability to capture complex relationships. As a case study, industrial wastewater treatment processes were modeled through a custom KG schema. Entities, relationships, and properties were represented as nodes and edges, while predictive models were incorporated to integrate process knowledge with data-driven insights. The population of the KG was performed manually to ensure semantic accuracy and consistency. The implementation demonstrates how KGs can unify heterogeneous data, support advanced querying through Cypher, and provide a foundation for predictive analytics. The results confirm the potential of KGs to enhance knowledge representation and decision-making in industrial domains, while future work may explore automation and scalability.