A Unified Framework of Temporal Information Expression in Geosciences Knowledge System
Shu Wang,Yunqiang Zhu,Yanmin Qi,Zhiwei Hou,Kai Sun,Weirong Li,Lei Hu,Jie Yang,Hairong Lv
DOI: https://doi.org/10.1016/j.gsf.2022.101465
IF: 8.9
2023-01-01
Geoscience Frontiers
Abstract:Time is an essential reference system for recording objects, events, and processes in the field of geosciences. There are currently various time references, such as solar calendar, geological time, and regional calendar, to represent the knowledge in different domains and regions, which subsequently entails a time conversion process required to interpret temporal information under different time references. However, the current time conversion method is limited by the application scope of existing time ontologies (e.g., “Jurassic” is a period in geological ontology, but a point value in calendar ontology) and the reliance on experience in conversion processes. These issues restrict accurate and efficient calculation of temporal information across different time references. To address these issues, this paper proposes a Unified Time Framework (UTF) in the geosciences knowledge system. According to a systematic time element parsing from massive time references, the proposed UTF designs an independent time root node to get rid of irrelevant nodes when accessing different time types and to adapt to the time expression of different geoscience disciplines. Furthermore, this UTF carries out several designs: to ensure the accuracy of time expressions by designing quantitative relationship definitions; to enable time calculations across different time elements by designing unified time nodes and structures, and to link to the required external ontologies by designing adequate interfaces. By comparing the time conversion methods, the experiment proves the UTF greatly supports accurate and efficient calculation of temporal information across different time references in SPARQL queries. Moreover, it shows a higher and more stable performance of temporal information queries than the time conversion method. With the advent of the Big Data era in the geosciences, the UTF can be used more widely to discover new geosciences knowledge across different time references.