A Multilayered Approach to Constructing a Patent Field of Technology Ontology
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Conventional classification systems struggle to represent the rapid advancement of technology and the increasing prevalence of interdisciplinary research, necessitating more dynamic and comprehensive approaches to organizing technological knowledge. This article presents a methodology for constructing a multilayered Field of Technology (FOT) ontology that explicitly separates a static layer and a dynamic layer. The static layer is grounded in the International Patent Classification (IPC) and enriched with Wikipedia, providing a stable foundation of core technological domains. For the dynamic layer, we derive technical concepts from English-language patent titles in the Google Patents Public Data using a custom NER model (PatentNER), and organize them through a hierarchical construction method that integrates density-based and hierarchical clustering. The resulting ontology comprises a multi-level hierarchy in which more than half of the dynamic concepts span multiple technical domains. To the best of our knowledge, this is the first study to construct such a large-scale, hierarchical FOT ontology directly from patent data, providing a scalable and adaptable resource for understanding the complex and interdisciplinary nature of modern technologies.
Fields of Study & Subjects
Semantic Web and Ontologies
Intellectual Property and Patents
Biomedical Text Mining and Ontologies
Publication details
- Journal / source
- ACM Transactions on Knowledge Discovery from Data
- Volume
- 20
- Issue
- 5
- Pages
- 1–37
- Publication date
- 27 March 2026
- Language
- EN