Semantic MediaWiki/Overall goals
The purpose of this page is to delineate the scope of the Semantic MediaWiki project more sharply. Of course, these goals are subject to discussion, but it is hoped that we do generally agree on most points. These major objectives affect the more fine-grained list of requirements.
What we want
First and foremost: The project is concrete and goal-directed. Its overall purpose is to obtain an implemented extension to MediaWiki that fulfills most of the needs for semantic annotation in various Wikimedia projects. It is believed that this can be done in a unified way, with a single common approach that ensures compatibility between the projects, and that keeps authoring as easy as ever. In addition, this should be achieved soon.
In any case, we want to keep the current content and structure of the Wikipedia as it is! We only allow users to make some of this content more explicit, such that it can be read out automatically, to be used and processed in software. This will typically involve some tagging or classifying of parts of the current content, but this must affect the normal editing experience only marginally.
Secondary goals of the project are to promote the development of internal and external applications for this technology and to explore the enormous new possibilities made possible by a machine-accessible Wikipedia. See the section on envisaged applications for a first impression.
What we do not want
- Create a complicated semantically-enhanced Wikipedia, where annotations can only be edited by ontology experts, or where annotation data makes up major parts of the articles.
- Turn Wikipedia into a playground or an experiment for ontologists.
- Change the scope of Wikipedia or alter its "encyclopaedic" character. (Content will not be new, but there will be better access to it!)
- Introduce heavy-load internal search engines and the like, that consume bandwidth, computing power, and programming effort. (In fact, we hope that our extensions allow for easy outsourcing of projects. Semantic annotations are only part of the data and they can be cached easily. We expect that some applications could even work on offline copies of some small part of the data, without sacrificing functionality or performance.)