Semantic MediaWiki/Envisaged applications

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

Here is the place to articulate your wildest fantasies on what could be done with all this semantic data!

Inside the Wiki[edit]

  • Powerful Searching: all the weird queries stated above should work (given that someone writes a user interface to ask such queries!)
  • Most of the functionality of today's Personendaten/Wikidata, with the exception of some envisioned features of Wikidata (like customized editing forms or predefined data-classes in MediaWiki). Full Wikidata functionality is somewhat orthogonal: it is more tailored toward scenarios that require uniform and rich input data (e.g. Wiktionary), while Semantic MediaWiki aims at more loosely structured settings (e.g. Wikipedia). However, the searching and sorting functions should be fully available very early.
  • Dynamically created pages: abolish plain article listings and overly specific categories and create consistent lists and galleries automatically. Even the articles themselves might be partly dynamic: place semantic queries after introductory text and let the query be replaced by its result as soon as the article is viewed.
  • New browsing functions: search queries that are simple can be executed quickly. Links on/below article pages can refer to search results, so as to provide a kind of "find similar topics"-link.
  • Use Wikipedia and Wikimedia Commons as a free image database ("I want a picture with a child and a cat on it").
  • Richly structured Wikis, including multiple perspectives, multiple levels of abstraction, dependency/contingency relationships, etc.
  • Easy visualization of content structure (categories, taxonomies, semantic nets, etc.). Direct editing of content structure.
  • Wiki content linked to dynamic models, simulations, visualizations. The Wiki content itself might provide data for the models.
  • Supports richer user access/rights models, including reputation systems.
  • Rich user relationships to the project and each other using semantic syntax on User pages and shared watchlists.

Offline analysis[edit]

  • Translate semantic data by inter-language links and thus compare Wikipedias of any languages.
  • Check for possible errors or inconsistencies to improve the quality of Wikipedia.
  • Create complex statistics from Wikipedia dumps.

External applications[edit]

  • Enable specialized desktop apps to query Wikipedia's knowledge selectively (or to ship with some frozen part of it):
    • teaching applications (e.g. a quiz game on geographic data)
    • media players (gather things like discographies from Wikipedia, along the lines of what amaroK starts doing today with XML-data from Amazon)
    • scientific applications (anybody needs a huge classification of species?)
    • and much much more
  • External web services:
    • sophisticated or specialized Wikipedia search engines (e.g. social networks based on rich user pages and shared watchlists).
    • question answering based on Wikipedia (e.g. integrated in major web searching engines)
    • Automatic detection of engineering data and specifications requiring review during ECP of interrelated process chains and component hiearchies and dependencies for distributed CAD/CAE/CIM connected supply/customer chains. (i.e. computer coordinated distributed concurrent engineering and manufacturing)
    • Integration of plain Wikimedia/Wikipedia data into all kinds of external applications to provide basic data. For example insurance companies may import Wikipedia data to evaluate insurance risk.
    • Integration with map data like Google Earth.
    • Making logical mind maps for people, example: if you are interested in Bach you may also be interested in Mozart, perhaps even Vienna and classical German literature.
    • Search engine crawling. Search engines will be able crawl the internet in much more sophisticated ways and automatically make connections.
  • Replacement of ontologies for annotating content:
    • Annotate any content by using the Wiki concepts as values for metadata in external applications. Based on the semantic relations between Wiki concepts, inference can be done.
    • The semantical relations should be possible to export in RDF (or, e.g., SKOS) format.

Artificial Intelligence - The Singularity[edit]

The full adoption of the Semantic Web with most of all available data converted may lead to something close to Artificial Intelligence (AI). It has been suggested that the arrival of AI will actually not be real intelligence but rather an advanced computational system. When IBM's Big Blue won the chess game against Kasparov it was not due to true artificial intelligence but rather smart computing of knowledge. The Semantic Web could allow computers to work with all of human knowledge in much the same way Big Blue worked with our knowledge of chess - and come up with marvelous things, beating human kind at our own game of knowlege. The technological singularity may arrive at the point when this process starts feeding on itself towards infinity. Wikipedia may very well be at the very forefront of this development.

This plan for a machine future is of course not embraced nor considered inevitable by all. It is increasingly critiqued by growing numbers including luminaries such as famed computer scientist Bill Joy and others suggesting the need for alternative plans giving more uplifting alternatives.

One example for such an alternative might be Project Halo, that is aiming towards the creation of a digital aristotle whom could serve as a tutor or scientific research assistent. Part of the project was the development of the Halo Extension for the Semantic MediaWiki, to optimize its ease of use and provide tools to reuse collaboratively authored knowledge for advanced question answering.

In the forefront of the transition to the Semantic Web[edit]

  • The transtion of the internet to the Semantic Web is a large undertaking that will allow for enormous productivity gains for mankind. The Semantic Media/Wiki will be in the forefront of this development with an early working model on a major data set. As the world's attention turns to the Semantic Web there will be increased focus on the Wikimedia and Wikipedia projects.

It is at present time impossible to fully envisage the full impact of the transition to the Semantic Web, other than to say that the potential is dramatic.


Semantic MediaWiki helping social networking on PeopleWiki[edit]

At Wikimania 2007, I will present a project whose goal is to implement a particular type of MediaWiki, which i call PeopleWiki.

In a few words : The main difference between a PeopleWiki and common wikis is that community members are themselves a subject of the information contained on the wiki: more than a user page, they are (through their vitae, hobbies etc..) the subjects of a wiki page. Thus PeopleWiki are both "normal wikis" when they deal with concepts, history etc... and "special" when they deal with community members. Basic ideas for regulating contributions and technical specifications are suggested in the paper and poster..

As I'm pretty new to MediaWiki, I do not know of any similar project, but it is likely someone already tried this idea. Anyway, we will try it out from next academic year on.

Now, - last but not least - the Envisaged applications about Semantic MediaWiki is: design/implement/test Ontologies for social networking --Ofol 23:55, 6 July 2007 (UTC)

Project Semantic MediaWiki.
This article is associated with the project Semantic MediaWiki.