Databases Reference
In-Depth Information
flights websites. But, using our example search engine, she is suggested convenient
flight connections between the two destinations, because it contains a large amount of
data so as to cover all the airports. She is also o
ff
ered convenient combinations of planes,
trains and buses. The provision of such information also necessitates the presence of a
large amount of internal as well as externals links between the datasets so as to provide
a fine grained search for flights between specific places.
Definition 10 (Relevancy).
Relevancy refers to the provision of information which is
in accordance with the task at hand and important to the users' query.
Metrics.
Relevancy is highly context dependent and can be measured by using meta-
information attributes for assessing whether the content is relevant for a particular task.
Additionally, retrieval of relevant documents can be performed using a combination of
hyperlink analysis and information retrieval methods.
Example.
When a user is looking for flights between any two cities, only relevant in-
formation i.e. start and end times, duration and cost per person should be provided. If a
lot of irrelevant data is included in the spatial data, e.g. post o
ces, trees etc. (present
in LinkedGeoData), query performance can decrease. The user may thus get lost in the
silos of information and may not be able to browse through it e
ciently to get only
what she requires.
Trust Dimensions.
Trust dimensions are those that focus on the trustworthiness of
the dataset. There are five dimensions that are part of this group, namely,
provenance
,
verifiability
,
believability
,
reputation
and
licensing
.
Definition 11 (Provenance).
Provenance refers to the contextual metadata that fo-
cuses on how to represent, manage and use information about the origin of the source.
Provenance helps to describe entities to enable trust, assess authenticity and allow re-
producibility.
Metrics.
Provenance can be measured by analyzing the metadata associated with the
source. This provenance information can in turn be used to assess the trustworthiness,
reliability and credibility of a data source, an entity, a publishers or even individual
RDF statements. There exists an inter-dependancy between the data provider and the
data itself. On the one hand, data is likely to be accepted as true if it is provided by a
trustworthy provider. On the other hand, the data provider is trustworthy if it provides
true data. Thus, both can be checked to measure the trustworthiness.
Example.
The example flight search engine constitutes information from several air-
line providers. In order to verify the reliability of these di
erent airline data providers,
provenance information from the aggregators can be analyzed and re-used so as enable
users of the flight search engine to trust the authenticity of the data provided to them.
ff
Definition 12 (Verifiability).
Verifiability refers to the degree by which a data con-
sumer can assess the correctness of a dataset and as a consequence its trustworthiness.
Metrics.
Verifiability can be measured either by an unbiased third party, if the dataset
itself points to the source or by the presence of a digital signature.
