Semantic Web

The semantic web represents a further development of the World Wide Web and Web 2.0, in which not only information is linked, but the meaning of things can be processed. It is also referred to as Web 3.0. In order to get closer to the goal of a semantic web, tools are being employed to convey the meaning of content to machines. In this context, Google is working on developing a semantic search. A first step in this direction has already been taken with the Hummingbird Update.

Background

Machines, that is computer programs, can only process the information with which they were previously fed or for which the human being has previously set up rules for processing. A computer differs from a person in that a person can also recognize the meaning of things. The semantic web is now intended to allow machines to automatically relate content and data based on algorithms in order to determine the meaning of the individual contents.

An example of the requirements for a semantic web is the web search. Today it is difficult to get a meaningful answer to a question such as “Why don’t bananas grow in the Alps?” if a website cannot be found that includes this exact word combination. If the semantic web works, such questions should be easy to answer by machines.

Implementation

In order to make the Internet “smarter,” people must help the machines describe the information in more detail. The W3C already provides standards for this: http://www.w3.org/standards/semanticweb/. Markups or standards for meta data will assist with this. They are briefly sketched out here:

• RDF (Resource Description Framework): This is used to describe the information of a website. RDFa also serves as a supplement, which allows RDF to be inserted into XML documents.
• Dublin Core: This is a standard for storing meta data on the content of a digital document in machine-readable form.
• RIF: This is used to define rules which are intended to establish the meaning correlations.

Practical examples

The semantic web is only in its beginnings. However, the development of search engines shows how much machines can now recognize the meaning of content.

One example is the Google Images Search, which can already recognize similar images with uploaded files. Google’s Knowledge Graph is also a result of the ever-improving semantic search. For example, it recognizes if a user wants to make a comparison.

The development of the Semantic Web is also a learning process. Many Google Group strategies show how this learning progress can be achieved through data mining. In addition, features such as the web History of users help the search engines to determine whether a user is looking for “Golf” the car model items related to the sport when entering the word of the same name into the search mask.