Abstract
Essential information on the early stages of a mission design is contained in Engineering Models. Yet, these models are often uneasy to visualise, query, let alone compare. This study demonstrates how Knowledge Graphs can overcome these data silos, interconnect information, provide a big-picture perspective, and infer new knowledge that would have remained hidden otherwise. Following the migration of CubeSats Engineering Models to a Knowledge Graph, two case studies are explored. The first case study illustrates how graph inference can derive implicit knowledge from existing explicit concepts. In the second case study, a Natural Language Processing layer is adjoined to the Knowledge Graph to enhances the analysis of textual content. The Natural Language Processing layer relies on the document embedding method doc2ve
Original language | English |
---|---|
Number of pages | 7 |
Publication status | Published - 30 Sept 2020 |
Event | 9th International Systems & Concurrent Engineering for Space Applications Conference (SECESA 2020) - Digital Duration: 30 Sept 2020 → 2 Oct 2020 https://atpi.eventsair.com/QuickEventWebsitePortal/20c06-secesa/secesa |
Conference
Conference | 9th International Systems & Concurrent Engineering for Space Applications Conference (SECESA 2020) |
---|---|
Period | 30/09/20 → 2/10/20 |
Internet address |
Keywords
- knowledge graph
- Machine Learning
- doc2vec
- space systems design
- concurrent engineering
- engineering models