Database Reference
In-Depth Information
The retrieved metadata, based on
RSVs
evaluated by the chosen matching function can be ranked in
decreasing order with respect to the user query, thus avoiding empty answers and suggesting an access
order to the geodata.
In the example proposed, by using the
close
matching function, both metadata are retrieved: Metadata
2 has a membership degree “1” being situated in late Summer, while Metadata 1 is also retrieved since
it partially satisfies the query condition “
close
”, thus meaning that it is associated to an observation of
glacier melting that is in proximity of the required temporal range 'late Summer'.
CONCLUSION
The proposal in this chapter originated within the context of a recent experience in European projects
where we had to cope with the actual job of both filling metadata of our georesources and training other
colleagues to create their own ones in order to create a discovery service for a European SDI. In carry-
ing out this activity we encountered several difficulties derived mainly by the constraints on metadata
type and formats imposed by the current INSPIRE implementation rules.
In fact, it happens very often that information on the geodata is not well known or completely trusted,
or it is even lacking. In some cases, the available geodata is produced manually by distinct experts in
charge of performing land surveys. Only successively this information is transformed into electronic
format, so that there can be a substantial temporal gap between the time of geodata creation and their
metadating, time in which pieces of information can get lost. Such imperfect information on metadata
call for new data models capable to represent and manage imprecision and uncertainty of the metadata
values. This aspect is particularly evident as far as the temporal metadata used to describe geodata, and
this is the aspect that has been analyzed in this chapter. We observed in particular that temporal resolu-
tion, time series and imperfect and/or linguistic temporal values are missing in the current INSPIRE
Implementing Rules, so we proposed distinct metadata elements to represent them and proposed a fuzzy
database framework to allow the representation of imperfect metadata and their flexible querying in
catalog services.
ACKNOWLEDGMENT
Two out of five authors of this paper carried out this research under temporary contracts. We wish to
acknowledge European Commission funding for this possibility, without which our work couldn't be
performed.
REFERENCES
Allen, J. F. (1983). Maintaining knowledge about temporal intervals.
Communications of the ACM
,
832-843. doi:10.1145/182.358434
Bordogna, G., Carrara, P., Pagani, M., Pepe, M., & Rampini, A. (2009). Extending INSPIRE metadata to
imperfect temporal descriptions, In the
Proceedings of the Global Spatial Data Infrastructures Confer-
ence (GSDI11), June 15-19 2009, Rotterdam (NL),
CD Proceedings, Ref 235.
