Database queries often include explicit or implicit temporal reference. Explicit queries ask e.g. for information in a given year or time period, and implicit queries e.g. look for information about changes over time like growth in a certain area.
Our datacenter nature and landscape (DNL) is basically a relational database with thematic data combined with a geographic information system (GIS) holding spatial data. The DNL has been semantically enriched with a bilingual eco- ontology and corresponding algorithms (semantic search, reasoning, consistency checking). The main purpose of the ontology is to allow open and intuitive search facilities for non-expert users. However so far there is no semantic support for temporal queries included yet.
Spatio-thematic information is stored in an OWL/DL ontology. Therefore an approach for temporal information is searched which fits seamless in the existing environment. Temporal information is available in the DNL database, however it is often given in a very coarse granularity (months or years), sometimes it is entirely missing. As a consequence the challenge of semantic support for temporal queries in the DNL, also includes handling vague and incomplete temporal information.
The work presented here, starts from typical temporal user queries (which were collected in use cases) and checks different approaches for their usability in our real-world eco-database DNL. It is found, that a subset of Allen's interval algebra relations, which actually corresponds to RCC-5 relations best fits Our needs. It is planned to be integrated into the existing thematic ontology in order to allow spatio-temporal and thematic queries.