Wikulu - Self-Organizing Wikis

Project Goals

The importance of web-based collaboration systems called Wikis has grown tremendously over the last years, e.g. Wikipedia, corporate wikis. As the usability of a wiki is initially very high, the amount of content grows very fast. A common drawback of wikis is however that the usability decreases with the increased content amount. The Wikulu - Self-Organizing Wikis project at the UKP Lab employs the latest Natural Language Processing (NLP) technologies to manage unstructured information, i.e. to structure the content in corporate Wikis. The objective of the project is thus to implement intelligent approaches to assist the user while creating, editing, or searching content. Wikulu should relieve the user of manual information management, leaving more room for productive work. Why is it called Wikulu? Kukulu is Hawaiian for to organize!

Project Publications

Counting What Counts: Decompounding for Keyphrase Extraction

Author Nicolai Erbs, Pedro Bispo Santos, Torsten Zesch, Iryna Gurevych
Date July 2015
Kind Inproceedings
PublisherAssociation for Computational Linguistics
AddressBeijing, China
Book titleProceedings of the ACL 2015 Workshop on Novel Computational Approaches to Keyphrase Extraction
LocationBeijing, China
Research Areas Ubiquitous Knowledge Processing, UKP_p_WIKULU, UKP_p_DKPro, UKP_a_NLP4Wikis, UKP_reviewed, UKP_s_DKPro_Core
Abstract A core assumption of keyphrase extraction is that a concept is more important if it is mentioned more often in a document. Especially in languages like German that form large noun compounds, frequency counts might be misleading as concepts “hidden” in compounds are not counted. We hypothesize that using decompounding before counting term frequencies may lead to better keyphrase extraction. We identified two effects of decompounding: (i) enhanced frequency counts, and (ii) more keyphrase candidates. We created two German evaluation datasets to test our hypothesis and analyzed the effect of additional decompounding for keyphrase extraction.
Full paper (pdf)
[Export this entry to BibTeX]

Important Copyright Notice:

The documents contained in these directories are included by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.


We are always looking for students who are interested in Wikulu and want to help us with our programming and research tasks. Please contact us if you want to know more!

Related Projects

The Wikulu project builds upon cutting-edge fundamental NLP technologies developed at UKP Lab to solve real-life knowledge management problems. It builds upon several successful projects ongoing at the UKP Lab, such as:

  • WiWeb funded by the Förderinitiative Interdisziplinäre Forschung: Utilizing Web Knowledge: Language Technologies and Psychological Processes
  • SIR 1+2 funded by the German Research Foundation (DFG): Extracting structured lexical semantic knowledge from wiki-based web 2.0 sources such as Wikipedia and Wiktionary and integrating contextually-aware semantic relatedness into information retrieval and keyphrase extraction
  • DKPro funded by UIMA 2007 Innovation Award and by two UIA 2008 Innovation Awards from IBM: Integrating NLP components in a repository of semantic information management software based on an industrial strength IBM’s Unstructured Information Management Architecture (UIMA) framework


The Wikulu - Self-Organizing Wikis project is funded by the Klaus Tschira Foundation.


A A A | Drucken Print | Impressum Impressum | Sitemap Sitemap | Suche Search | Kontakt Contact | Webseitenanalyse: Mehr Informationen
zum Seitenanfangzum Seitenanfang