Semantic Web and IoT
9
2.1.6 Earth Observation
Combining semantics with deep learning techniques into Earth Observation data has
been a hot topic during the past few years. The primary reason behind this tendency is
that while Earth Observation data are increasing rapidly, semantics offer an intelligent
fusion between data deriving from heterogeneous sources and high-level solutions
in decision-making issues, as they enhance knowledge discovery. In most cases,
the problems are associated with physical resources management, environmental
protection and monitoring. For this purpose, several ontologies and systems have
been developed in literature. Some of them are presented below.
Most systems combine semantics in order to achieve knowledge discovery by
fusing data from heterogeneous sources [31–35] while others are responsible for
semantic querying for image retrieval [36]. The main challenge encompasses, have
to do with combining data from heterogeneous sources (e.g. OpenStreetMaps [32,
34], satellite and aerial images [32], Google Earth images [34]) which is achieved
using different fusing techniques (like FuseNet architecture [32]). The scope of using
semantics in earth observation data has to do with semantic labeling in most cases
[32, 33].
The ontologies developed to map earth observation data are associated with hydrological data [37] and environmental monitoring data [35]. More specifically, the
ontology presented in [37] represents sensors, observation and hydrological events
classes, while Modular Environmental Monitoring ontology (MEMOn) [35] builds
a more extensive structure which contains a large amount of aspects like disaster,
temporal, environmental material, sensor, environmental process, geospatial, observation and measurement and infrastructure modules.
Intelligent Interactive Image Knowledge Retrieval (I
3 K R) [31] is a system which
conducts semantic-based Knowledge discovery by using EO data archives. The system uses a hybrid ontology approach to interconnect data from different sources and
DL reasoning services to apply semantic restrictions. PREDICAT [35] is a system
which aims in interconnecting data from heterogeneous monitoring systems using
ontologies, data integration and reasoning techniques to produce knowledge from
existing natural disasters and predict possible future natural catastrophes.
2.1.7 Creative Industries
According to the Department of Culture, Media and Sport, there is a plethora of
creative sectors that are identified as belonging to the creative industries [38]. This
subchapter encompasses state-of-the-art related work and applications from all sectors except for those related to Cultural Heritage which are presented thoroughly in
the following subsection.
There are great opportunities in creative industries for administering digital content. In that aspect, an ontology-based framework named as V4Ann was developed
as part of the V4Design platform [39]. Its main purpose was the knowledge representation, the semantic aggregation from multiple sources and the combination of
9
2.1.6 Earth Observation
Combining semantics with deep learning techniques into Earth Observation data has
been a hot topic during the past few years. The primary reason behind this tendency is
that while Earth Observation data are increasing rapidly, semantics offer an intelligent
fusion between data deriving from heterogeneous sources and high-level solutions
in decision-making issues, as they enhance knowledge discovery. In most cases,
the problems are associated with physical resources management, environmental
protection and monitoring. For this purpose, several ontologies and systems have
been developed in literature. Some of them are presented below.
Most systems combine semantics in order to achieve knowledge discovery by
fusing data from heterogeneous sources [31–35] while others are responsible for
semantic querying for image retrieval [36]. The main challenge encompasses, have
to do with combining data from heterogeneous sources (e.g. OpenStreetMaps [32,
34], satellite and aerial images [32], Google Earth images [34]) which is achieved
using different fusing techniques (like FuseNet architecture [32]). The scope of using
semantics in earth observation data has to do with semantic labeling in most cases
[32, 33].
The ontologies developed to map earth observation data are associated with hydrological data [37] and environmental monitoring data [35]. More specifically, the
ontology presented in [37] represents sensors, observation and hydrological events
classes, while Modular Environmental Monitoring ontology (MEMOn) [35] builds
a more extensive structure which contains a large amount of aspects like disaster,
temporal, environmental material, sensor, environmental process, geospatial, observation and measurement and infrastructure modules.
Intelligent Interactive Image Knowledge Retrieval (I
3 K R) [31] is a system which
conducts semantic-based Knowledge discovery by using EO data archives. The system uses a hybrid ontology approach to interconnect data from different sources and
DL reasoning services to apply semantic restrictions. PREDICAT [35] is a system
which aims in interconnecting data from heterogeneous monitoring systems using
ontologies, data integration and reasoning techniques to produce knowledge from
existing natural disasters and predict possible future natural catastrophes.
2.1.7 Creative Industries
According to the Department of Culture, Media and Sport, there is a plethora of
creative sectors that are identified as belonging to the creative industries [38]. This
subchapter encompasses state-of-the-art related work and applications from all sectors except for those related to Cultural Heritage which are presented thoroughly in
the following subsection.
There are great opportunities in creative industries for administering digital content. In that aspect, an ontology-based framework named as V4Ann was developed
as part of the V4Design platform [39]. Its main purpose was the knowledge representation, the semantic aggregation from multiple sources and the combination of