Visit our archive

Búsqueda semántica

Hearing that SEO is dead is a statement that makes me smile thinking that you’ll be in shock when you see him walking!

It is something that is in constant change and its evolution can not be more alive. With the new devices, searching systems, voice recognition, personal assistants, SIRIs, Cortanas, GoogleNow and more, search results in something more human and more personal. It’s not just typing on a screen like a robot “Italian restaurant madrid” now we also demand “I want pizza dinner in Madrid”, we seek more semantic form.

But what is the semantic search? Well, it should be clarified that semantics is the part of linguistics that studies the meaning of the expressions so is the interpretation of linguistic signs. Thus, semantic is trying to go beyond the significant meaning and approach.

Semantics studies the meaning of what is said

Therefore, the semantic search seeks to improve searching accuracy by understanding user intent and the contextual meaning of the terms used in the search.

This means that words simply fail to be words to become concepts (with a range of connotations that give a real sense) and this is where comes in disambiguation and the quid of the whole matter. For a human it is relatively easy to interpret the phrase “OMG, how smart you are!” Depending on the environment, situation and context, but for a robot (searcher) is not that easy: is it praising me? is insulting me ?, is it telling me yes or not? ?, are you hungry ?, are you thirsty ?, what I wanted to say is…

Mainly due to the large number of synonyms and multiple meanings that the language contains, its especially important to apply semantic search systems in search engines. Very roughly, this process of semantic search is basically divided into the following aspects:

  1. User interpret the question extracting the most relevant concepts of the sentence.
  2. Use this set of concepts to create a query and use it against ontology (formal definition of types, properties, and relationships between entities) system.
  3. Present the results to the user.

To run this, a searcher -Google for example- have to understand two fundamental concepts: intention and context. To do this we use method called Latent Semantic Indexing (LSI), whose key feature is its ability to extract the conceptual content of a document, establishing partnerships between those terms that occur in similar contexts, and identify relationships between the terms contained use in a document.

In addition, for some time now and increasingly insistently, Google has been offering incentives to developers of web pages to add a title, description, headings, long text, formatting and so on, to better position and, in turn, go “learning” about what exactly what is said on the web.

Joint projects of Google with Bing and Yahoo! as (where Yandex also joined), intended to add tags on your pages in order to help search engines better understand web sites, that is, a code capable of providing semantics to HTML.

In this way, the search is becoming “wise”, “human” and “semantic” because these data we deliver consciously (and others perhaps not so consciously) is able to assimilate and understand. What kind of data we deliver unaware? There are many more than you think: Search History (global and user), user location, vocabulary, spelling, synonyms, variations …

Whenever we do a search, we paged or not, we click or not, look again or not … the search registers, study and learn to improve the next time you’re at that request and thus know how to act to provide a better result: more relevant, more effective and more “human”. “Search engines are not what people think but how people think: drive and response.” (Incidentally, if you liked this quote, I recommend a fantastic and intriguing film Alex Garland released a year ago, “Ex Machina“).

So, how do I do SEO semantic? Being more natural and more human. We repeated ad nauseam that Google gives increasing importance to natural content, quality and enriched. Thinking and anticipating the user questions to answer what is sought on the web. Writing good and fluid, with varying synonyms, plurals, conjugations, adjectives, related terms, etc. Logically and always focused to the user. Stay with two concepts:

  • Co-occurrence: close relationship of two or more terms in a unit of text.
  • Co-citation: when content makes a reference to a second and a third, the latter two are related to each other and also with the first.

With all this, it is clear that current search engines, led by Google, have not done more than scratching the surface of this vast field of semantic search, but it is also true that a scratch if it It leaves a scar deep enough, and they are many and varied efforts that are being putting in this regard. Moreover, the advantages and possibilities of this new generation of searcher engines are endless and exciting.

I leave you with a video explaining the Semantic Web and its applications by Metaweb (by the way, already acquired by Google in 2010).

It is so cool, isn´t it?

¡Esperamos tus comentarios!


Uso de cookies

Este sitio web utiliza cookies para que usted tenga la mejor experiencia de usuario. Si continúa navegando está dando su consentimiento para la aceptación de las mencionadas cookies y la aceptación de nuestra política de cookies, pinche el enlace para mayor información. ACEPTAR