Automatic semantic modeling by cross-modal retrieval

Date

2022

Authors

Xu, R.
Mayer, W.
Wang, Y.
Zhang, H.Y.
Ning, D.
Duan, Y.
He, K.
Zaiwen, F.

Editors

Advisors

Journal Title

Journal ISSN

Volume Title

Type:

Conference paper

Citation

Proceedings - 24th IEEE International Conference on High Performance Computing and Communications, 8th IEEE International Conference on Data Science and Systems, 20th IEEE International Conference on Smart City and 8th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Application, HPCC/DSS/SmartCity/DependSys 2022, 2022, pp.2142-2150

Statement of Responsibility

Conference Name

24th IEEE International Conference on High Performance Computing and Communications, 8th IEEE International Conference on Data Science and Systems, 20th IEEE International Conference on Smart City and 8th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Application, HPCC/DSS/SmartCity/DependSys 2022 (18 Dec 2022 - 20 Dec 2022 : Chengdu)

Abstract

Semantic models of data sources describe the concepts and relations within the data. Building semantic models with the help of a common ontology is a key step in automatically publishing the semantics of structured data into knowledge graphs. However, modeling semantics of data manually requires considerable human cost, and expertise and can be error-prone. Most of the research contributions are focused on semantic annotations of source attributes, and yet it is significantly critical to explicitly infer the relations between attributes. In this paper, we present a novel approach that applies the cross-modal retrieval model as a scoring function to identify the most plausible semantic model for a target data source. As far as we know, this is the first technique that uses cross-modal retrieval models to explore the relations within data sources and semantic models in an end-to-end manner. Preliminary experiments demonstrate that our approach outperforms the state-of-the-art method

School/Discipline

Dissertation Note

Provenance

Description

Access Status

Rights

Copyright 2022 IEEE

License

Grant ID

Call number

Persistent link to this record