Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Constructing industrial knowledge graph through ontology and link prediction

Wan, Yuwei, Chen, Zheyuan, Liu, Ying ORCID:, Packianather, Michael ORCID: and Wang, Rui 2023. Constructing industrial knowledge graph through ontology and link prediction. Presented at: IEEE International Conference on Automation Science and Engineering (CASE), Auckland, New Zealand, 26-29 August 2023. 2023 IEEE 19th International Conference on Automation Science and Engineering (CASE). IEEE, pp. 1-6. 10.1109/CASE56687.2023.10260566

[thumbnail of Manuscript_CASE_2023_yuwei_V1.2.pdf]
PDF - Accepted Post-Print Version
Download (707kB) | Preview


Given that the numerous data embedded in manufacturing processes and products are separated, it is challenging to tackle and integrate heterogeneous data in industrial scenarios. In this context, an industrial knowledge graph (iKG) has been developed as a promising semantic organisation to leverage the rich information from multiple resources. However, relations are usually missing and hidden in original iKGs, which results in the necessity for iKG completion. Given these two perspectives, a framework of iKG construction is proposed based on ontology and link prediction in this study. Firstly, an ontology design framework is deployed to generate domain-centric ontologies after extracting numerous data (e.g., entities and relations). Secondly, the missing relations between each couple of entities are discovered over existing knowledge to increase the number of edges that complete and refine iKGs. Thirdly, iKG visualisation is conducted by importing data into the generated ontology. The feasibility and effectiveness of the proposed framework are substantiated and demonstrated in a case study using real-world data.

Item Type: Conference or Workshop Item (Paper)
Date Type: Publication
Status: Published
Schools: Engineering
Publisher: IEEE
ISBN: 9798350320701
ISSN: 2161-8070
Date of First Compliant Deposit: 25 May 2023
Last Modified: 06 Nov 2023 11:11

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics