Toward Spatio-Temporal Patterns
Our world is changing at an increasing pace causing the validity intervals of information to shrink. Therefore, the change of information becomes important information itself an important information resource that should be exploitable by query languages.
- PDF / 3,203,458 Bytes
- 25 Pages / 439 x 666 pts Page_size
- 91 Downloads / 200 Views
2.1 Introduction Our world is changing at an increasing pace causing the validity intervals of information to shrink. Therefore, the change of information becomes important information itself-an important information resource that should be exploitable by query languages. In particular, information about the position and extent of objects is of high interest in many areas of science and elsewhere (vehiele tracking, history, security, health care, ecology, weather prediction, urban planning, and so on). In all these areas relevant information has to be extracted from huge data sets. Beyond data mining that often serves the purpose of identifying interesting objects in spatial or spatiotemporal data sets, queries are of interest that find out about relationships between objects once they have been identified. In the realm of spatiotemporal databases this means to find out, in particular, ab out the changes that objects and their relationships undergo. Moreovcr, many applications rcquirc not only the detection of one single change, but rather look for sequences of changes describing particular developments. Existing spatiotemporal database systems and query languages offer only basic support to query changes of data. Most of these systems allow the formulation of qucries that ask for changes at particular time points. However, it is often very difficult, if not impossible, to express queries for sequences of such changes. In other words, existing query languages do not offer a systematic, scalable concept to query developments of objects and relationships; instead they require the user to encode these by a number of individual conditions. This method is awkward, in particular, because it also requires the formulation of additional side conditions. Moreover, this approach does not work in queries for arbitrary numbers of changes. A concept of spatiotemporal patterns and their integration into spatiotemporal query languages can help to elose this gap. Spatiotemporal patterns enable the formulation of queries about complex object developments; their integration into query languages allows the formulation of queries that are
R. de Caluwe et al. (eds.), Spatio-Temporal Databases © Springer-Verlag Berlin Heidelberg 2004
30
Martin Erwig
currently not possible. Moreover, spatiotemporal patterns can also simplify the formulation of queries that are possible with existing languages. Beyond the integration of spatiotemporal patterns into a textual query language we will also discuss the design of a visual query language and a corresponding user interface to support the formulation of spatiotemporal patterns and queries. This is of particular importance because most users of spatiotemporal data (for example, scientists) do not have a formal computer education and do not know how to use a query language, in particular, query languages for complex data like spatiotemporal data. Moreover, the fast growing set of available spatial and spatiotemporal data and its wide dissemination through the Internet increases the class of poss
Data Loading...
