Information Technology Reference
In-Depth Information
Pack(o
1
,DC
1
)
=
(o
1
,{t
1
,t
3
})
Pack(o
1
,DC
2
)
=
(o
1
,{t
1
,t
2
,t
4
})
Pack(o
1
,DC
3
)
=
(o
1
,{t
1
,t
2
,t
3
,t
4
,t
5
,t
6
})
We have to note at this point that there is not any qualitative difference between
DIPs and AIPs from our perspective. The only difference is that AIPs are formed
with respect to the profile decided for the archive, which we can reasonably assume
that it is usually richer than user profiles. For example in Fig.
8.9
three different
AIPs for module
o
1
are shown for three different DC Profiles. The DIPs of module
o
1
for the profiles
DC
1
,
DC
2
and
DC
3
are actually the corresponding AIPs without
the line that indicates the profile of each package.
We should also note that community knowledge evolves and consequently DC
profiles may evolve over time. In that case we can reconstruct the AIPs according
to the latest DC profiles. Such an example is illustrated in Fig.
8.10
. The left part
of the figure shows a DC profile over a dependency graph and at the right part it
is a newer, enriched version of the profile. As a consequence the new AIP will be
smaller than the original version.
8.2.3.4 Dependency Management and Ingestion Quality Control
The notion intelligibility gap allows for a reduction in the amount of dependen-
cies that have to be archived/delivered on the basis DC profiles. Another aspect
of the problem concerns the ingestion of information. Specifically, one question
which arises is whether we could provide a mechanism (during ingestion or cura-
tion) for identifying the
Representation Information
that is required or missing. This
requirement can be tackled in several ways:
Fig. 8.10
Revising AIPs after DC profile changes
