202
G. E. Modoni and M. Sacco
3.3.2 Security of the Graph
An increasing number of threats and vulnerabilities can jeopardise cybersecurity of
information systems such RDF stores. In particular, data confidentiality, integrity,
and availability (so called CIA) are at risk. Thus, it is essential to identify and apply
new security measures to mitigate the risks of attacks against the systems to an
accept-able level, thus protecting the information processed by those systems. One
crucial mechanism to handle security of data stored within an RDF store consists
in applying the access controls at graph level or even at lower triple level through a
more fine-grained security policy, thus allowing to grant users a granular data access
[25]. In this regard, RDF stores must provide proper tools to manage these privileges
to allow users to perform specific operations.
3.3.3 Versioning of the Graph
In any collaborative scenario, many different involved stakeholders can interact concurrently by applying continuous changes to the information concerning the ongoing
activities. Under these conditions, it is essential to have a mechanism capable to
version these information, thus allowing to ensure the consistency among different
versions of these persisted information. Specifically, in the context of the SWT, the
versioning capability allows to keep track of different versions of RDF graphs, which
are also defined in terms of complex hierarchies of other RDF graphs [25]. The latter
can be in turn versioned to handle all their evolutions. Moreover, since different versions of graphs can be potentially linked, the versioning should allow to specify how
these different versions are linked [18]. Finally, the versioning system should support
users to apply and manage changes, by allowing to compare different versions of
graphs, highlight the differences, and also merge the different users involved in the
design process [28].
3.3.4 Handling Multimodal Data in a Knowledge Repository
In various fields, information concerning the same phenomenon can be collected
through different modalities of acquisition by using specific equipment [20]. In fact,
due to the rich set of properties characterizing the real phenomena, it is difficult that a
single modality offers a holistic knowledge about the interesting phenomenons. The
growing availability of these different modalities referring to the same phenomenon
poses the need to properly manage and combine these multimodal data in the repositories where they are persisted [33]. Among the technical challenges to realize the
management of these multimodal data, there is its organization, link, and fusion of
the interdisciplinary information as within a knowledge repository [48]. Under these
conditions, as the information can be represented in this system under the form of
different data models, it is needed the reconciliation and mediation of these models.
In this regard, one potential solution can be based on the alignment of the different
G. E. Modoni and M. Sacco
3.3.2 Security of the Graph
An increasing number of threats and vulnerabilities can jeopardise cybersecurity of
information systems such RDF stores. In particular, data confidentiality, integrity,
and availability (so called CIA) are at risk. Thus, it is essential to identify and apply
new security measures to mitigate the risks of attacks against the systems to an
accept-able level, thus protecting the information processed by those systems. One
crucial mechanism to handle security of data stored within an RDF store consists
in applying the access controls at graph level or even at lower triple level through a
more fine-grained security policy, thus allowing to grant users a granular data access
[25]. In this regard, RDF stores must provide proper tools to manage these privileges
to allow users to perform specific operations.
3.3.3 Versioning of the Graph
In any collaborative scenario, many different involved stakeholders can interact concurrently by applying continuous changes to the information concerning the ongoing
activities. Under these conditions, it is essential to have a mechanism capable to
version these information, thus allowing to ensure the consistency among different
versions of these persisted information. Specifically, in the context of the SWT, the
versioning capability allows to keep track of different versions of RDF graphs, which
are also defined in terms of complex hierarchies of other RDF graphs [25]. The latter
can be in turn versioned to handle all their evolutions. Moreover, since different versions of graphs can be potentially linked, the versioning should allow to specify how
these different versions are linked [18]. Finally, the versioning system should support
users to apply and manage changes, by allowing to compare different versions of
graphs, highlight the differences, and also merge the different users involved in the
design process [28].
3.3.4 Handling Multimodal Data in a Knowledge Repository
In various fields, information concerning the same phenomenon can be collected
through different modalities of acquisition by using specific equipment [20]. In fact,
due to the rich set of properties characterizing the real phenomena, it is difficult that a
single modality offers a holistic knowledge about the interesting phenomenons. The
growing availability of these different modalities referring to the same phenomenon
poses the need to properly manage and combine these multimodal data in the repositories where they are persisted [33]. Among the technical challenges to realize the
management of these multimodal data, there is its organization, link, and fusion of
the interdisciplinary information as within a knowledge repository [48]. Under these
conditions, as the information can be represented in this system under the form of
different data models, it is needed the reconciliation and mediation of these models.
In this regard, one potential solution can be based on the alignment of the different