Typed link

A typed link in a hypertext system is a link to another document or part of a document that includes information about the character of the link. For example, rather than merely pointing to the existence of a document, a link might also specify that the document supports the conclusion of the article pointing to it, that it contradicts the article pointing to it, that it is an older version of the document, that it serves to define the word next to the link, that it is an index to other documents of the same type, or some other relationship. This allows a user to take actions such as searching only certain types of links or displaying them differently. It may also allow browsing software to do things like pre-fetching documents it expects the user to browse.

While not an important part of Internet-based hypertext systems, typed links were a common feature in pre-Internet hypertext systems such as Xanadu, NoteCards, HyperWriter, IBIS/gIBIS, and others. Although they can be very useful, the lack of a standardized set of link attributes such as "Supports Position"/"Refutes Position" as well as the difficulty of applying the attributes has always hindered the use of typed links beyond prototyping purposes. Any system of attributes would turn a hypertext corpus into a semantic web, and permit more sophisticated processing.

Version 4 of the HTML standard defined by the World Wide Web Consortium supports typed links using the rel (forward relationship) and rev (reverse relationship) attributes. These attributes are applied to either the <link> tag (for links between whole documents) or the <a> tag (for links from a specific part of a document). For example, the tag <link rel="contents" href="top.html"> specifies that the document "top.html" is a table of contents for the work that includes the document you are currently reading, and the tag <link rel="next" href="chap3.html"> specifies that "chap3.html" is the next document in logical sequence after the one you are reading.

The following link types are defined in the HTML 4 standard:

• alternate
• stylesheet
• start
• next
• prev

• contents
• index
• glossary
• copyright
• chapter

• section
• subsection
• appendix
• help
• bookmark

Unofficial link types also exist. For example, Google has proposed to reduce the effect of comment spam in blogs with "rel=nofollow".

Semantic link

A semantic link is a typed link where the element itself includes meaningful information about the link (semantics). The link describes some external relationship or issue. Such links are the basis of a semantic network.

Examples:

• In XML, a BookTitle element such as <BookTitle isbn="0812511816" />. You can use the semantic information provided in the element (the ISBN) to create a link from this element to an online bookseller.
• Any syntax that can express genealogy, such as "A is-mother-of B". Other relations like "B is-child-of A" may be implied by it.

The Semantic Web as described by Tim Berners-Lee is one example of the many attempts over many decades to define standards for semantics of links.

Among the earliest were those used by early hypertext systems in the 1980s such as KMS, NoteCards and IBIS.

HTML included rel and rev attributes in its hyperlinks to store additional information about the nature of the links. There was at least one attempt (in 1994) to define a set of standard rel and rev attributes to attach to HTML anchor tags to make them into typed links that would express semantic properties of the link. Although this IETF RFC failed, it set the standard for many other attempts to follow. Over the next ten years, a great many attempts to set standards for particular subsets of the World Wide Web rose and fell.

These attributes of the HTML links are used by some Microformats to add meaning to relationships. XHTML Friends Network (XFN), for instance, uses the "rel" attribute to allow web authors to identify their relationship to people they link from their websites. Other Elemental Microformats use these too - Vote Links use rev rev="vote-for", rev="vote-against" or rev="vote-abstain"; Tags use rel="tag", Directory uses rel="directory".

The Semantic Link Network Model

Semantic link is a link with a semantic property (factors, tags or a complex description), The following are simple examples of semantic links: A-isPartOf->B, A-isSouthOf->B, A-isFriendOf->B, and A-{studentof, coauthorOf}->B. Isolated definition of semantic link is meaningless. It should be defined as a part of semantic link network.

The Semantic Link Network seems similar to the semantic network, hyperlink network, and RDF, but they are different in goal, architecture, technology and research method.

Semantic Link Network (SLN) consists of semantic nodes, semantic links and reasoning rules. The semantic nodes can be any resources, classes of resources, or even a semantic link network. Semantic links can be established by tools or automatic discovery approaches. The reasoning rules are for semantic reasoning. New semantic links may be derived out by rule reasoning. The semantics of a semantic link network will change if its rules are changed.

Semantic Link Network is a description of relations among objective existences rather than to represent fine human knowledge. It pursues semantic richness rather than correctness.

Due to the semantic richness of semantic nodes, automatic discovery of semantic links is feasible, especially, when relevant metadata is available. This topic is never studied in traditional semantic network.

SLN naturally supports relational reasoning, analogical reasoning, and inductive reasoning. SLN would be enriched after reasoning. SLN is statically defined but autonomously evolved, it can be localized or decentralized. SLN is not a simple extension from the hyperlink to the semantic link, it is a self-organized semantics-rich data model for managing various resources. Compared with other approaches to the Semantic Web, SLN better inherits the characteristics of the Web. SLN naturally supports relational query.

With the addition and removal of semantic links, a semantic link network may evolve into semantic communities. The approach to discovering semantic communities in a large semantic link network is different from those for general graph. The semantic communities dynamically support intelligent applications. Further more, the semantic link network has distinguished network effect, which is useful in social network study [H.Zhuge, Communities and Emerging Semantics in Semantic Link Network: Discovery and Learning, IEEE Transactions on Knowledge and Data Engineering, 21(6)(2009)785-799].

The Semantic Link is further extended to complex link and super link.

External links

• "Link relationships". HTML 4.01 Specification. W3C. http://www.w3.org/TR/html4/struct/links.html#h-12.1.2. 

• {{cite web|url=http://www.knowledgegrid.net/~H.Zhuge/SLN.htm