submitted to the Developers Track of WWW2007 in Banff, Alberta, Canada
Introduction
In this short paper we will demonstrate how embedded RDFa in Weblogs can be used as a medium for data-transfer between desktops. A combination of two existing Semantic Web tools - the desktop-based Semantic Blog authoring tool semiBlog and the Semantic Clipboard application - allows one user to export and blog data from various desktop applications such as electronic addressbooks, calendars or bibliographic databases, and another user to import the same data back into their own applications.
Both tools have been presented independently before; and while both have a different focus, they complement each very well in the data-exchange scenario described in this paper. The motivation behind semiBlog [Möller et al. 2006] is the reuse of data from other desktop applications for the annotation of blog posts with formal metadata, i.e. Semantic Blogging. The focus of semiBlog is on providing a simple, unobtrusive user interface for the semantic annotation of blog posts, using techniques such as drag&drop (from other applications into semiBlog) and autocompletion (using a system-wide index such as Spotlight™). The idea behind the Semantic Clipboard [Reif et al. 2006] on the other hand is the exact opposite - instead of providing functionality to lift data from the desktop to the (Semantic) Web, it detects Semantic metadata in web pages and allows to import it back into various desktop applications.
A Scenario - The Web as a Universal Clipboard
Fig. 1 - Using the Web as a Clipboard
Together, both applications allow to use the Web as a universal clipboard, where semiBlog allows to copy data onto it, and the Semantic Clipboard allows to paste it back. This is illustrated by the scenario depicted in Fig. 1. A blogger called Tom announces an upcoming meeting in company-internal blog, which is used to disseminate information within his team. He uses semiBlog to write the post and annotate it with data about the meeting from his calendar, and with the contact details about Marco, who is a visitor and will be attending the meeting as a guest. During the upload to the web, the annotations are converted into RDFa ([Note1]) and embedded into the post. Jerry, who is in Tom's team, reads the blog post, and uses the Semantic Clipboard application to paste the metadata contained in the post back into the relevant applications on his desktop.
Architectures
Both semiBlog and the Semantic Clipboard application have similar architectures, even though both operate in opposite directions. Both applications enable the integration of arbitrary ontologies and applications by using a plugin-based architecture, and define interfaces that describe each plugin in terms of a pairing a desktop application with an appropriate ontology (e.g. the Outlook application with the iCalendar ontology).
Conclusion
We presented a solution for the transfer of structured data between various desktop applications over the (Semantic) Web. Our approach combines two existing prototypes. Both have been developed independently, but really are two sides of the same coin, and illustrate how RDF can be used for the integration of structured data.
References
[Möller et al. 2006] Knud Möller, Uldis Bojars and John G. Breslin. Using Semantics to Enhance the Blogging Experience. In 3rd European Semantic Web Conference (ESWC2006), Budva, Montenegro, June 2006. (pdf)
[Reif et al. 2006] Gerald Reif, Martin Morger and Harald Gall. Semantic Clipboard - Semantically Enriched Data Exchange Between Desktop Applications. In SemDesk2006 Workshop at the 5th International Semantic Web Conference (ISWC2006), Athens, GA, USA, November 2006. (pdf)
[Note1] Both semiBlog and the Semantic Clipboard can use RDFa to export and import annotations, respectively. However, also other means of associating HTML documents to RDF graphs are supported, such as linking two documents.