TUD Logo

TUD Home » ... » Research Projects » Finished Reference Projects » Finished Reference Project

Computer Networks

Finished Reference Project


WebKnox - Web Knowledge Extraction

  • Internet Information Extraction
  • Information Quality
  • Internet Information Retrieval
  • Fact Extraction
  • Entity Extraction
  • Machine Learning


Information is distributed across the World Wide Web. It is often duplicated and partly even incorrect. In order to get an overview about a certain topic or entity (a movie for example), one often needs to manually crawl the Web and collect information in a time consuming manner. Recently, there is a trend towards aggregating and structuring data to make them accessible for human users and machines. Wikipedia, DBpedia, Freebase, TrueKnowledge and other knowledge bases have been developed following that goal. The project WebKnox (Web Knowledge Extraction) aims for an automatic knowledge base construction from the Web. WebKnox finds and extracts entities and facts from the Web, it assesses the quality of the extractions and stores them semantically. A semantic representation allows not only human users to access the knowledge but machines as well. This is important in helping the Semantic Web to progress and make the knowledge available as Linked Data.

» Projects homepage

Project leader:

Dipl.-Medieninf. David Urbansky


Recent publications:

Student Assignments:

Show all

Last modified: 20th Oct 2018, 6.54 AM
Author: Webmaster