DK1:WikiSense - Mining the Wiki

Title: WikiSense - Mining the Wiki (ID DK1)

  • Language: English/German? | License: ?
  • Room size: medium/large
  • Category: Research,Visualisation
  • type: presentation/workshop?
  • Budget requirements: {{{budgetrequire}}}
  • Budget priority: {{{budget}}}

Author[s]: Daniel Kinzler (aka de:Duesentrieb).

  • Contact: daniel@brightbyte.de [OTRS]
  • Contacted by: ?
  • (From: | available days: ?)
Abstract

THERE IS AN additional ABSTRACT IN OTRS AS PDF I would like to give a lecture on a project for mining the wikipedias in order to create a semantic network and dictionary. I'm Duesentrieb at the de:wp, the commons and on meta, Gearloose at the en:wp.

There is a loose collection of ideas at http://brightbyte.de/wiki/?WikiWordExtraction (german) that outline the project. I will prepare the lecture and paper in english.

Here are a few key points:

  • analysis is focused on link structure, analysis of natural language is avoided.
  • isolation of concepts and terms that refer to the same concept in different languages.
  • identification of groups of synonyms, homonyms, hyponyms
  • extraction of sematic relations like is-a, part-of and other ontological concepts

This project is (hopefully) going to be my major thesis (Diplomarbeit) at the Universität Leipzig. I am planning to do this in cooperation with the Wortschatz project (http://wortschatz.uni-leipzig.de). Hopefully until august I will have a first prototype running and will be able to show some results.

I know of at least two people who are planning on giving related lectures at wikimania: nsh (of en:wp) and GerardM (of nl:wp). I'm in contact with them and will try to coordinate and cooperate with them in preparation of out talks at wikimania. I CC this mail to them for teir information.

I would like to do a full lecture on this, but if the schedule is tight, I can go for a short presentation, or alternatively would share time with nsh and/or gerard (I have not asked them about this, though). Please tell me, so I can plan ahead and know how much material I should prepare.

About the author[s]:


Status information in the templates is not up to date. Please see Internal:Speakers/Categories for final status information.


  • accept: JV, EM (Daniel is a smart guy, this should be interesting), PD
  • reject:
  • status:
  • comments: not sure I see the interest in this, but I don't object if other people want to accept it. Angela