This task will focus on the aggregation and transformation of integrated linked geospatial data into search-optimized data structures such as optimized index structures. These data structures can then be used by applications to run enhanced queries focussing on topics and customer motives on the integrated GKS data. Motive and topic based search is a new search paradigm, combining elements of information retrieval and questions answering. Such searches are particularly important in the e-Commerce domain, where customers search for parameters of products and services. An example of a motive and topic based search is “winter holiday with culture and mountains”, which could return travel options offering hiking as well as cultural highlights during the winter school holidays in the region of the searcher. Such searches require substantial (spatial) background knowledge as it can be only provided by the Web of Data. The current state of the art is focused on querying data statically. Only by using the Linked Data paradigm it is possible to derive data from the Data graph dynamically. However, this approach would cause a large effort if the data is not organized well (state explosion during derivation). Therefore, the research task is located on the research area where ontology-based question answering overlaps with heuristic-based derivation as well as statistic-based IR.
Other Tasks in this Workpackage
|T6.1||Customer data selection, retrieval and preparation|
|T6.2||Design integration methods and develop prototype to utilize RTD results|
|T6.4||Evaluation and testing of the search prototype|