The current practice in the creation of eGovernment websites is often unsatisfactory for both the maintainer and the end user. Public authorities commonly use content management systems or portals which require arranging the individual services “by hand”, whereas end users frequently fail to discover these services for lack of suitable navigation possibilities.
Semagogue was designed in the course of a diploma thesis at SAP Research, Karlsruhe, and also partly implemented as a prototype software. It aims to improve website generation and maintainability, especially in the eGovernment sector. It focuses on semi-automatic generation of pages and navigation structures from semantic metadata.
An important method to create such metadata is faceted (multi-dimensional) classification. It means that the items to be made available online are classified not just using a single, rigid taxonomy, but any number of "facets" or "dimensions" that allow both clasification and website navigation from multiple points of view. Proven examples of such facets in the public sector are subject, audience, or life situation.
If such a faceted classification is consequently employed, it gives the user more and better possibilities to discover the services they need. It also lets them explore the website more easily, in the common case that they only have a vague idea of what they are looking for.
The following images show screenshots from the classification software, and a web browser presenting a page of the generated website. A fictitious form for a "student loan application" is classified and made available online.
Additional Information & Evaluation Survey
I have put up a website at http://www.till.krullmann.de/semagogue, which contains a more detailed description of how the software works, several screenshots and two demonstration screen videos, everything in both English and German.
The website also contains an evaluation survey where you can provide feedback about this prototype or the concepts behind it. Such feedback is very valuable to decide whether and how research in this direction will continue.
