Moving from traditional data maintenance to a future of linked, multi-temporal, self-improving data with multiple representations and feature-based metadata that is analysis ready.
The Challenge
The Victorian Government are responsible for providing authoritative foundational spatial data (FSD i.e. Vicmap), which involves labour intensive manual update and maintenance of disparate datasets. Key challenges include managing frequent updates, including from space-based sources; managing multiple versions, including authoritative and machine-learned data; and managing data provenance, including per-feature rather than per-layer metadata. The plethora of new space-based data streams and new technological capabilities associated with Artificial Intelligence (AI) and cloud computing provide an opportunity for fast, effective, and efficient updates to our critical digital infrastructure, such as Vicmap FSD. This project investigated the use of space-based data streams and other data sources for updating and enriching Victoria’s FSD.
Partners
This Proof of Concept (PoC) project, led by FrontierSI, was a collaboration with RMIT University for the Victorian Department of Transport and Planning, funded by SmartSat CRC.
The Solution
The Dynamic Vicmap project pioneered a fundamental shift from static, layer-based data management to dynamic, integrated databases supported by semantic, knowledge-based data management. The approach is founded on internationally recognised standards and ontologies (including GeoSPARQL, PROV-O, and the Australian FSDF). Using the latest graph database technologies enabled the construction of transparent and integrated knowledge graphs through semantic enrichment of authoritative FSD. As a proof-of-concept, the Dynamic Vicmap project demonstrated a semantic, knowledge-based workflow for managing and updating Victorian FSD, with a focus on Vicmap Hydro. The demonstrator automates the process of semantic enrichment, delivering integration, interoperability, and dynamic updates of different authoritative and non-authoritative Vicmap hydro, property, and flood data sets. For example, the demonstrator illustrates how easily water features with multiple different representations can be queried; or how simply all Victorian properties affected by flooding can be identified.
Impact
The PoC showcased the opportunity and provides a framework to move from traditional data maintenance to a future of linked, multi-temporal, self-improving data with multiple representations and feature-based metadata that is analysis-ready. The linked data and semantic enrichment approach improves decision making by enabling intelligent search, retrieval, querying and analysis of data stored with multiple representations, reducing the steps to insights. The PoC also facilitated better understanding of the technical implementation of the knowledge graph-based approach including the importance of ontology engineering. It is important to continue to increase awareness of the benefits of using knowledge graphs for maintaining and updating FSD.
Contact
To learn more, contact FrontierSI at contact@frontiersi.com.au.
