| Social bookmarking sites generally organize their | | | | classification of Internet resources (such as web sites) |
| content using tags. Social bookmarking sites are an | | | | is done by human beings, who understand the content |
| increasingly popular way to locate, classify, rank, and | | | | of the resource, as opposed to software which |
| share Internet resources through the practice of | | | | algorithmically attempts to determine the meaning of a |
| tagging and inferences drawn from grouping and | | | | resource. This provides for semantically classified tags, |
| analysis of tags. | | | | which are hard to find with contemporary search |
| History | | | | engines. |
| The concept of shared online bookmarking dates back | | | | Additionally, as people bookmark resources that they |
| to April 1996 with the launch of Within the next three | | | | find useful, resources that are of more use are |
| years online bookmark services became competitive, | | | | bookmarked by more users. Thus, such a system will |
| with venture-backed companies like Backflip, Blink, | | | | "rank" a resource based on its perceived utility. This is |
| Clip2, Hotlinks, Quiver, and others entering the market. | | | | arguably a more useful metric for end users than other |
| Lacking viable models for making money, most of this | | | | systems which rank resources based on the number |
| early generation of social bookmarking companies | | | | of external links pointing to it. |
| failed as the dot-com bubble burst. The contemporary | | | | Automatic Notification |
| concepts of social bookmarking and tagging took root | | | | Since the classification and ranking of resources is a |
| with the launch of the web site oneview in 1999[1] and | | | | continuously evolving process, many social |
| del.icio.us, in September of 2003. | | | | bookmarking services allow users to subscribe to |
| Functional Overview | | | | syndication feeds (see RSS) based on tags, or |
| In a social bookmarking system, users store lists of | | | | collection of tag terms. This allows subscribers to |
| Internet resources, which they find useful. Often, these | | | | become aware of new resources for a given topic, as |
| lists are publicly accessible, and other people with | | | | they are noted, tagged, and classified by other users. |
| similar interests can view the links by category, tags, or | | | | Disadvantages |
| even randomly. Some social bookmarking systems | | | | There are drawbacks to such tag-based systems as |
| allow for privacy on a per-bookmark basis. | | | | well: no standard set of keywords (also known as |
| They also categorize their resources by the use of | | | | controlled vocabulary), no standard for the structure of |
| informally assigned, user-defined keywords or tags | | | | such tags (e.g. singular vs. plural, capitalization, etc.), |
| (see folksonomy). Most social bookmarking services | | | | mistagging due to spelling errors, tags that can have |
| allow users to search for bookmarks which are | | | | more than one meaning, unclear tags due to synonym |
| associated with given "tags", and rank the resources | | | | antonym confusion, highly unorthodox and |
| by the number of users which have bookmarked | | | | "personalized" tag schemas from some users, and no |
| them. Many social bookmarking services also have | | | | mechanism for users to indicate hierarchical |
| implemented algorithms to draw inferences from the | | | | relationships between tags (e.g. a site might be labeled |
| tag keywords that are assigned to resources by | | | | as both cheese and cheddar, with no mechanism that |
| examining the clustering of particular keywords, and | | | | might indicate that cheddar is a refinement or |
| the relation of keywords to one another. | | | | sub-class of cheese). |
| Advantages | | | | The separate (but related) tagging and social |
| This system has several advantages over traditional | | | | bookmarking services are, however, evolving rapidly, |
| automated resource location and classification | | | | and these shortcomings could possibly be addressed |
| software, such as search engine spiders. All tag-based | | | | in the near future. |