What is it about?
This paper aims to address the limitation of describing the evolution process of complex systems (CS) through knowledge graphs (KGs). In general, the asymmetric changes caused by updated knowledge make the entities challenging to align, thereby hindering the effective analysis of CS evolution. In this context, we propose a Siamese-based Graph Convolutional Network (GCN) model, called SiG, to bridge the evolved domain knowledge of the current transportation system (TS) across three intergenerational processes, namely partial, high, and full autonomy. The SiG tries to provide fundamental support for analyzing the future-stage development of autonomous TS (ATS).
Featured Image
Photo by Edz Norton on Unsplash
Why is it important?
The research work is important for addressing unresolved issues in terms of low accuracy, efficiency, and effectiveness in aligning asymmetric KGs. To achieve this, we design a full-process SiG to generate evolved KGs, optimize entire-graph computation, enhance feature extraction, and optimize model training. The evaluation results experimentally suggest that the proposed model has average improvements of 23.90% and 37.89% in accuracy and efficiency, respectively. The SiG can, consequently, provide significant implications for TS evolution analysis and offer a novel perspective for research on CS limited by continuously updated knowledge.
Perspectives
Read the Original
This page is a summary of: SiG: A Siamese-based Graph Convolutional Network to Align Knowledge in Autonomous Transportation Systems, ACM Transactions on Intelligent Systems and Technology, February 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3643861.
You can read the full text:
Contributors
The following have contributed to this page