Workpackage description: 12.4: Large Scale Resource Discovery and Presentation Demonstrator

Workpackage number: 12.4

Start date or starting event: Month 11

Lead Partner: ILRT (1)

Participant short name: ILRT W3C-ERCIM CCLRC HP STILO
Participant number: 1 2 3 4 5
Person-months per participant: 6 0 0 0 0

Total number of deliverables: 1


Description of work

A major application of the Semantic Web is to enable better discovery of information on the (Semantic) Web - to enable more relevant information to be found more quickly. This application has strongly influenced the design of Semantic Web technologies such as the Dublin Core Metadata Element Set (DCMES) and RDF. This workpackage will investigate how simple searches by end-users of large scale Semantic Web data can be implemented and deployed. Since the results of such searches can return a large and almost arbitrary graph of information, it will also investigate techniques to turn this complex graph into something that can be presented to the user in an appropriate manner for their device. This could include HTML for desktops, other formats for palmtops and mobile devices.

Specific areas of work include:

Large Scale Resource Discovery

Investigate using large scale full text searching of literal data along with querying of semi-structured Semantic Web data in order to provide higher quality retrieval while still delivering a simple interface to the end user. This will involve using Semantic Web storage and query systems as well as integrating structured querying with full text approaches. Searching the Semantic Web and retrieving results will involve information about a variety of subjects, stored in different representational models and potentially containing only partially-understood concepts. The latter is a crucial and potentially very common occurrence in Semantic Web systems since not everything about the data schema will be known in advance. This will have to be considered so that search systems can work both cross-schemas when the RDF Schema or other mapping is present and when such information is not available or missing.

Presentation of Semantic Web data

The data from a search may need to be shown in multiple forms - as one large result set and as individual parts. The form of the presentation to the end user will have to be easy to understand, appropriate for type of data and potentially flexible enough to allow multiple views of the information. In addition, with an interface generated from multiple sources it is important for the user to understand the source, or provenance, of the information in order to judge the quality. This might be provided by quality marks provided by Web Of Trust machinery or other authentication method. The provenance has to be clearly presented to end users, for example on the web, by familiar design metaphors such as context, icons or (like a portal). For small devices such as palmtops and mobiles, other appropriate interfaces should be possible.