Back to top

TACKO - Tag-based Context Dependent Knowledge Organisation

Last modified Oct 6

Objective

The long-term objective of the project is the development of a new kind of organization structure that combines the advantages of classical hierarchical folder structures and tagging systems while mitigating their respective shortcomings. Particularly it aims to

  • improve the presentation of the contents of tagging systems to allow new modes of navigation
  • support the user in classifying information consistently by offering suggestions for labels
  • provide user interfaces and storage mechanisms that allow the user to dynamically reorganize information

The focus is neither on personal information management nor on collaborative information organization. Instead, the goal is to develop an organization structure that is as broadly applicable as folder hierarchies or tags.

Current Activities

Recently, several extensions to the commercial enterprise collaboration platform Tricia were made that resulted from this project:

  • A faceted search interface was developed to browse the contents of wikis. With this extension, users do not have to create hierarchical menu structures to browse wiki contents, but the tags assigned to these pages are used to provide faceted navigation options. This way, a list of pages can for instance be filtered by year as well as by project name given that such tags are used in the wiki.

  • An interface for the consolidation of the tags currently used in the system was developed. It allows the dynamic analysis of tagged contents and enables the user to add tags to or remove tags from large numbers of documents. It is currently in experimental state and not part of the official releases of Tricia.
  • In the context of the Hybrid Wikis project, tags play a major role in structuring the contents of wikis. Particular features are developed to be applied in this project.

Background

With the growing amount and the dynamics of digital information that has to be managed nowadays, the limits of classical systems for information organization, like hierarchical folder structures, become apparent. Fast changes regarding the subjects of stored information items and unanticipated information needs require organization systems that can dynamically be adapted to new kinds of content -- with respect to the information subjects -- and classification schemes respectively.

In the past years, tags (i.e., freely-chosen text labels assigned to information items by users) appeared on the web as a means to flexibly classify arbitrary information items. While there are some success stories, tags are currently only used to a limited extent since with the flexibility come new challenges: It is hard to maintain consistency regarding the usage of terms in an information base since the meaning of a tag is usually not defined. Synonyms and homonyms are thus inherently problematic. Together with the fact that relations between tags are not explicitly defined (e.g., category-subcategory relations) it is hard to browse the contents of tagging systems.

From our point of view, it is particularly desirable that new organization systems still offer modes of browsing and navigation that are familiar to the user. In addition to full-text search, the contents should be made accessible based on the user defined categories in ways that exploit the implicit relations between categories and go far beyond what we know from so-called tag clouds.

Related student projects