Social tagging

from Wikipedia, the free encyclopedia
Extract of a tag cloud from the Netbib internet diary

Social tagging or collaborative tagging describes the common keywording (indexing), in which users assign descriptors (keywords, tags ) to content with social software without predefined rules . A so-created collection of keywords is folksonomy (folksonomy) called. Tags can be visualized as a tag cloud (word cloud ) in which the most popular keywords are displayed at their largest.


The folksonomy is mainly used on websites or in the communities they offer to tag their content. Other users can then find this information by searching for a keyword. Popular items indexed in this way by many people are blog entries , photos, or social bookmarks . The users act in open communities without defined indexing rules. The joint indexing serves mainly the indexing .

Origin and origin

The origin of the suitcase word Folksonomy from “folk taxonomies”, ie lay taxonomies , is traced back to Thomas Vander Wal . Folksonomy was first used in 2003 on the delicious website .

Jon Udell stated in 2004 that this type of indexing was already known, but the possibility of feedback from individual users was new .

Isabella Peters presented a comprehensive study on folksonomies with regard to knowledge representation and information retrieval in 2009.

Advantages over controlled development

Through a folksonomy, each user can contribute something to the keywording. On the one hand, the categorization effort is distributed over many shoulders; on the other hand, better search results are achieved if the information objects are categorized by those who also use them. Due to the mostly large number of users, information and relationships that the individual did not notice should be made visible.

In addition to the individual benefit for self-organization, the individual user has the option of making their keyword collection accessible to the general public. For example, documents with identical keywords or users with similar interests (identified by their keywords) can be associated. The open sharing of the keyword assignments of individuals with others offers the community a good search option (joint development of an information space), but also allows individual users to become aware of other objects or other perspectives by assigning the keywords to users.

Disadvantages compared to controlled development

With conventional manual indexing, for example by librarians , mostly classifications or other centrally managed controlled vocabularies are used. In the Folksonomy, on the other hand, there is no instance that determines which keywords are to be used and which are not. Hence the categories are fragmented; so z. B. the same thing is referred to by some users in the singular (example: book ) and by others in the plural ( books ). In addition, there may be keywords in several languages ​​( book , book , livre , etc.). In the case of compound terms, if only one word and not several are technically permitted, you can choose between different separators (open_access, open-access) or combined words (openaccess) .

Another disadvantage of free indexing arises from synonyms and homonyms : Words can stand for completely different concepts, and the exact meaning of a keyword can often only be clearly understood in context. For example, the English word “apple” generally stands for the fruit apple, but the companies Apple and Apple Records can also be meant. If a user searches for photos of a summit meeting with the politicians Kohl and Bush on a hosting website such as Flickr with the keywords Kohl , Bush , Gipfel , photos of plants ( cabbage and bush ) and (mountain) peaks could also be found.

To counteract this problem, related keywords can be displayed and methods of semi-automatic indexing (suggesting suitable tags when assigning tags) can be used. A further corrective is the mass of users that could induce users to orientate themselves on the most popular catchphrase. Data mining methods such as clustering can also be used. This allows groups of similar resources to be distinguished from one another. In addition, several keywords should always be assigned to describe the content.

Collaborative indexing with controlled vocabulary

New internet projects combine collaborative indexing with lexical or semantic databases, such as Wikipedia or the semantic DBpedia , to offer a special form of controlled vocabulary .

See also


Web links

Individual evidence

  1. Isabella Peters: Folksonomies. Indexing and Retrieval in Web 2.0 . De Gruyter, Berlin 2009, ISBN 978-3-598-25179-5
  2. Faviki
  3. Semantic Tagging with Faviki - ReadWriteWeb
  4. Zigtag ( Memento of 7 December 2010 at the Internet Archive )