Speaker
Description
Ontology technology plays a vital role in information management within the fields of architecture, engineering, and construction. AI-driven automated and semi-automated ontology construction has emerged as a key research direction, significantly improving the efficiency and scalability of ontology construction while addressing the limitations of traditional methods that rely heavily on domain expert knowledge and manual input. Urban data formats have evolved from single-text sources to multimodal data, demonstrating rapid growth. However, current research on automated ontology construction predominantly focuses on single-source data, particularly text-based data, with limited studies on the fusion and automatic construction of ontologies from multimodal data.
To address this gap, this study investigates AI-based methods and technical pathways for automated ontology construction driven by multimodal data. First, a systematic literature review was conducted to summarize typical technical methods used at key stages of the ontology automation process, including concept extraction, relation identification, and hierarchical structure construction. The study further analyzes the performance of machine learning algorithms, such as large language models, reinforcement learning, and graph neural networks, at these stages. Second, knowledge extraction and multimodal data fusion strategies are explored for various data types, such as text, images, audio, and structured tables. Finally, this study proposes an AI-driven automated ontology construction method leveraging multimodal data, using a smart factory as a case study to demonstrate the use of various data types, including sensor readings, images, video streams, and operation logs, and to evaluate its practical feasibility.
The proposed method improves the efficiency of knowledge representation and the accuracy of semantic expression, contributing to the theoretical advancement of ontology construction. Moreover, its practical feasibility and scalability highlight its potential for broad application in intelligent systems, such as smart cities and smart buildings, thereby supporting enhanced ontology-based knowledge management and informed decision-making.
| Keywords | Automated Ontology Construction; Multimodal Data; AI-Driven Approach; Smart Manufacturing，Knowledge Representation |
|---|---|
| Best Congress Paper Award | Yes |
