What are entities in NLP and how do they improve SEO?
The entities in NLP are a fundamental
component of the new revolution of the SEO. Instead of looking only at
keywords, search engines are beginning to understand natural language. This
implies that we have to rethink our strategies in order to continue positioning
in search engines. So let's start at the beginning: we tell you all the
secrets of the entities .
What are entities and what are they for?
Entities are a concept that Google has been using since 2012,
but lately they have become more relevant thanks to its new algorithm, Google
Bert, which focuses on natural language processing or NLP .
Before, SEO was oriented around keywords and was not able to
distinguish the intention of the user. For example, it was possible that
"Italian restaurant in Madrid" and "Italian food in Madrid"
would return different search results, since Google's algorithm was not able to
identify that the user's intention was the same in both cases.
Now, instead, the analysis focuses on the entities of the
text. An entity is a named "thing", such as a famous person, a
landmark, an object, and so on.
Entities can be of two categories: proper nouns that are assigned to unique entities (for example, a person or a specific place) and common nouns. According to Google, it is good practice to consider that if something is a noun, it can be classified as an entity.
Another key aspect of NLP is that Google is now able to
establish the relationship between entities , usually through
sentences with a subject, a verb and a predicate. For example, you may
know that the entity "Dalí" is related to the entity
"Barcelona". All this information is used to interpret the
intention of users when searching and giving appropriate answers, regardless of
the specific words that a user uses. Thus, Google Bert would be able to
identify that "Italian restaurant in Madrid" and "Italian food
in Madrid" refer to the same entity and would offer the same results for
both searches.
Another very relevant concept to understand entities in NLP
is co-occurrence. The more frequently two entities are mentioned together,
and the more authoritative the result that mentions them, the stronger the
relationship between them.
Parameters that make up an entity
·
NAME: the name that represents the entity.
·
TYPE: the type of entity (in the next section we
explain the classification used by Google).
·
METADATA: For most entity types, the Wikipedia URL and
Knowledge Graph MID are included here.
·
SALIENCE - A score that defines the importance of an
entity within a document. Scores close to 0 indicate less importance, and
the closer the score is to 1, the more relevant that entity is.
·
MENTIONS: mentions to an entity in a document.
·
SENTIMENT: the opinion expressed regarding the entity
in the text. It can go from -1 (negative opinion) to 1 (positive opinion).
Types of entities according to Google
·
UNKNOWN - Unknown entity type
·
PERSON: person
·
LOCATION: location
·
ORGANIZATION: organization
·
EVENT: event
·
WORK_OF_ART: artwork (paintings, movies,
songs ...)
·
CONSUMER_GOOD: consumer product
·
PHONE_NUMBER: phone number
·
ADDRESS: address
·
DATE: date
·
NUMBER: number
·
PRICE: price
·
OTHER: another type of entity
How to
optimize your website with entities
Optimizing websites for Google Bert means
taking your mind off keywords and thinking about entities . The
idea is to write natural texts that respond to users' questions. In
addition, there are two very important issues that we must take into account:
how to disambiguate the entities that appear in our texts and how to make
Google consider our brand or company as an entity.
How to disambiguate entities in NLP
One problem with entities is that in many
cases the search terms that a user enters can be ambiguous; For example,
if you write "Barcelona", it can refer to the city or the soccer
team. To make Google correctly index our entities, we have to
"help" it understand what we are talking about. Sign up these
recommendations to get it:
·
Analyze
the entities that appear in your text. If before SEO optimization started
from the analysis of keywords, now it must start from the analysis of the
entities that appear in your text and the relationships between them that we
want Google to understand.
·
Use
structured data. Schema.org is a
data markup that is defined and that Google is able to understand, therefore,
whenever possible, mark the data with schema so that Google goes directly to
that source of information.
·
Take
advantage of the co-occurrence. For
Google it is essential to understand how the different entities are related to
each other. To make you understand that two entities are related, we will
have to use the same terms in different content, with a similar frequency and
proximity.
·
Use links
to indicate relationships. Google
knows that when one website links to another, it is very likely that their
contents are related. You can take advantage of this both when designing
the internal link structure of your page and in your link building strategy .
How to make Google consider you an entity
Making Google consider your brand or your
company as an entity helps you improve visibility and positioning, increase
your authority and control the information it shows about you. These tips
will help you increase the chances that Google will classify you as an entity:
·
Have your own website and a verified social
media presence.
·
Have the Organization schema.org data
markup with the Same As attribute referenced to your social networks.
·
Have an account on Google My
Business .
·
Have your own Wikipedia entry.
·
Having inbound links from authoritative
pages that are thematically related to the entity you want to position.
· Have content that talks about your products and services and that fits thematically with the semantic category assigned by Google to your website.
READ MORE:-