A Multi-modal Spatio-temporal Data Semantic Association Model Based on Ontology

Abstract

Multi-modal spatio-temporal data is mixed data with spatio-temporal information, various types and complex modalities. Different modal data are usually related to each other. Effective organization and associated spatio-temporal data are the focus of massive data management and data information mining. Aiming at the characteristics of multidimensional, multi-scale, multi-temporal and multi-modality of spatio-temporal data, by analyzing its rich spatio-temporal information and semantic information, this paper proposes a multi-modal spatio-temporal data semantic association model with time information, spatial information and content object semantic information as three related factors. At the same time, based on the characteristics of multi-modal spatio-temporal data features low-level features heterogeneity, the ontology model of spatio-temporal data semantic expression that can describe data information uniformly is defined and constructed based on Ontology theory. The association query is realized by constructing multiple instances of different modal spatio-temporal data. The experimental analysis shows that the multi-modal spatio-temporal data semantic association model can effectively correlate spatio-temporal data of different modalities and has certain scalability.