Semantically Enhanced Catalogue Search Model for Remotely Sensed Imagery

Ya Lin,Hao Xu,Yuqi Bai

DOI: https://doi.org/10.1109/jstars.2016.2590835

IF: 4.715

2016-01-01

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

Abstract:Keyword-based search enabled by catalogue services is now the dominant way to query remotely sensed imagery. One of its major limitations is that searchable attributes have to be maintained in the underlying metadata database. This study investigates the feasibility of mediating semantic query and catalogue search together to allow more searchable parameters without any changes to the existing metadata database and catalogue service. Limitations of a catalogue's textual search capabilities are analyzed. A use case of searching for sea ice imagery using search criteria that are absent in the NASA ECHO (the U.S. National Aeronautics and Space Administration EOS Clearing House) catalogue service is presented. An ontology dedicated for remotely sensed sea ice data collections is introduced. Details of a two-step hybrid metadata search model, i.e., collection-level discovery search enabled by ontology query, and granule-level inventory search fulfilled by catalogue service, are presented and evaluated. Our results show that this semantically enhanced catalogue search model could easily extend the existing catalogue service to allow more searchable parameters, and, at the same time, maintain a backward compatibility with them. The lessons learned may be useful to others' modeling of characteristics associated with geoscience data collections, and thereby providing enhanced geoscience data search capabilities.