Semantic Emergence From Social Tagging Systems

Social Tagging Systems (Folksonomies)

Folksonomy is a user-generated classification system of web contents. It is also known as collaborative tagging and social tagging. Folksonomy-base systems allow users to tag web resources (videos, images, links and etc.) with their chosen words or phrases. Social tagging systems consist of at least three set of elements: tagger, digital objects and tags. Taggers are the persons who interact within the closed community. Digital objects are the shared resources, and tags which are used for describing shared resources. (Thielen, et al, 2010)

Folksonomy term is coined by Thomas Vander Wal, derived from folk and taxonomy. Vander Wal identified two types of folksonomy: broad and narrow. In broad folksonomy (such as Delicious), many users tag particular content with their own vocabularies, thus creating a greater amount of metadata for that content. In narrow folksonomy (such as Flicker), few users tag an object with a limited number of terms which used to get back to the object. (Vander Wal, 2005)

• Definition 1:

• A folksonomy is a tuple F: = (U, T, R, Y) where:

• U is a finite set of users,

• T is a finite set of tags,

• R is a finite set of resources, and Y is a ternary relation between them, i. e., Y ⊆ U ×T × R, whose elements are called tag assignments.

Figure 1.

Model of tagging system (adapted from Weller et al, 2011)

Folksonomies are a cheap method of documents indexing and describing the masses of information on the web, it has proven a powerful alternative to existing top-down categorization techniques, such as taxonomies or predefined dictionaries due to its flexibility. (Wetzker et al, 2010) Folksonomy-based systems allow users to generate the content and classify it in their own way based on their individual own experience, linguistic styles and preferences without relying on a previously defined terms, which reduces the required cognitive efforts. On other hand, this freedom in the choice of vocabularies as tags leads to many problems such as ambiguity and lack of standardization. However, handling these problems will contribute much in improving information retrieval, recommendations and etc.