Semantic web: The real power of the Internet remains

2009/05/01 Leturia Azkarate, Igor - Informatikaria eta ikertzaileaElhuyar Hizkuntza eta Teknologia Iturria: Elhuyar aldizkaria

• Internet

Using all the information and internet options is not always easy. On the one hand, there is a huge amount of information, and to immerse ourselves without drowning in this flood of information, we need search engines like Google or Elebila. But with them, it is often not easy to find what we need. For example, when there are many results, it is not easy to distinguish between the possible meanings of the word we have given to look for, what are the pages about the meaning we want (when searching for the word "Texas" will appear, for example, pages about the state of the US, the musical group, the book or many other things at a time). Or many times there is no good way to differentiate between good and bad information, since the demonstration by the search engines in the first positions does not always ensure quality. On the other hand, in many cases it is not possible to unify all kinds of services, and when we want to buy something, for example, to compare prices and options, we only have to look at the web pages of different stores.

Source of problems, HTML format

All these problems come from the initial network design. The web is, in short, a collection of hypertexts located in the infrastructure of the Internet, that is, a collection of documents that refer to each other, and to encode the hypertext is used the HTML format created by Berners-Lee himself. The characteristics and limitations of this format are the cause of all problems.

HTML or HyperText Markup Language is a text marking language that explains how to describe a text or its parts. In HTML language, this descriptive markup is made using tags that are inserted between the symbols in the text itself. For example, in HTML to indicate that a part of text is a first-degree title, the h1 label ( h1 First Degree Title /h1 ) is used and to write in italics the em tag ( em italics text /em ). There are many labels of this type for links, images, etc. Browsers interpret this marking and show the user the page in a proper way.

The semantic web is a parallel network that would exist next to the HTML network that is encoded for people, a knowledge base understandable by machines. This will allow the effective processing and crossing of information. The semantic web incorporates intelligence to the Internet network.

12RF

But most HTML tags are to describe the structure and appearance of the text, so that a human being can properly view and understand the text. They are not designed to improve or facilitate the automatic treatment of machines. And search engines and other Internet tools are just machines. Only with HTML tags do not have enough information to work well and cannot understand the text as we do people. Thus, when it's only text and HTML tags, a search engine can't know if a page in which the word "Java" appears refers to the island or programming language, or if, on a page that talks about a product, people speak well or badly about it, or that a page is from an online store that sells a certain product...

The semantic web has solution

The creator of the World Wide Web himself is looking for a solution. In fact, the semantic web is one of the most recently worked guidelines by the international organization World Wide Web Consortium led by Tim Berners-Lee: a project that aims to correct the errors of the initial design.

The labeling of the HTML format is used to describe the documents and their relationships. The semantic web describes objects, people, etc. and their relationships. Instead of the labels reflecting the shape and structure of the page, the meaning of the elements of the sheet is labeled. In the semantic web, there could be labels for, for example, the statement and description of musical groups, others for the declaration and description of people, others for the description that some people form a musical group and so with all things.

In this way, search engines could differentiate from the pages containing the word "Scorpions" which are the corresponding to the musical group and which animals, and show the results grouped based on these possible meanings. Or create specialized search engines in musical groups. Or it could be detected with relative ease that the same product is sold in different online stores and build services that expose all price options together. Or, when detecting theatrical events and their location, we could automatically complete a calendar of all theatrical performances in a country. Or through the semantic labeling of scores given by users to a website or a product, search engines or stores could also rank by score. And the semantic web would have thousands of applications of this type that we still can't imagine.

Therefore, the semantic web would be a network parallel to the HTML network encoded for people, a knowledge base understandable by machines, encoded in expressive formats of semantic. Besides the web in natural language, we would have another structured text. This representation of knowledge could be understood by machines, treated effectively, inferred from new knowledge...

There is no doubt that the semantic web can be a great change in the Internet. But, how is it done to make power a reality? What tools are necessary? And what do we have?

It will continue...