Tip:
Highlight text to annotate it
X
Hello! I am Lucas from vetail-x.com
and today I'm gonna tell you
about the next big step forward
in the field of search engines.
At least since Google’s Hummingbird update
the term „semantic web“ is very popular.
So what’s that about?
Semantics deal with
the meaning of information in general
and with the meaning of words, in particular.
Let’s assume, for example, in the biography
of a person the word „Essen“ appears.
„Essen“ is the name of a German city
Does the person live there or was she born there?
But „Essen“ in German also means eating, so it could also be the person’s favorite pastime.
In case the place of residence is meant:
Who else lives there?
Where did the person live before?
If you answers all these questions
And connect the different meanings
And relations with each other
We develop a so-called semantic network.
Which basically is
a mindmap.
The objective of the semantic web
is to
Create such a semantic network
For all web content there is.
Why?
Because it structures the incredibly huge amount of data
In a very accessible way.
The most popular use case of semantic web
At the moment is probably Apple’s Siri
As well as Google’s „OK“, which is very similar.
When it comes to pure text
Google slowly catches up with Wolfram Alpha
the grand dame of semantic search
which has been delivering decent results
since 2005.
All of them aim
at answering actual questions
and delivering
important information.
Starting with very simple things
like recommendations what to buy
or the weather forecast that is displayed automatically
when planning a trip
as well as processing and classifying
technical texts.
All this becomes possible
with semantic web.
So if I book
a trip to the mountains
I get recommendations for hiking boots
potential weather warnings
and a list of the local
emergency numbers.
If the system believes that
I have a special interest in alpine flowers
I even get
a paper about
Cyclamens.
So, ideally the semantic web
actually delivers
only relevant information.
Actual websites
don’t appear in the results at all
or only in the form of excerpts.
For SEO as such
this is less than interesting
for the user on the other hand it is very comfortable.
So, how does this work?
How do the machines understand
what a website is about?
The answer is: Metadata, metadata, metadata.
Metadata!
Especially for Google
It is very important that
a website does not only contain information itself
but also information
about this information.
While in earlier times a few keywords were considered enough
now each proper name should ideally be
connected with additional information about it.
Furthermore, search engines expect
precise information about the
relations between the different parts of the content
on the page.
At the moment, there are different
approaches to this.
Some methods place the data directly in the code of a page
others use external files.
But an actual
universal standard for this procedure is still missing.
In the description of the video
you can find information about the most common methods
with details about their implementation.
It’s easy to see though
that it involves a lot of
effort to provide metadata
for all the web content there is,
and it’s questionable if this
will ever become reality.
But even if it becomes real
only for part of the internet
the benefit for us will be so great
that in a few years we will ask ourselves
how we ever could live
without the Semantic Web.
*Yeah!*
I’m Lucas from vetail-x.com.
*You’re not!*
And the guy behind the camera
is the most incompetent intern we ever had!
*Ok, comment section Kommentarleiste, blah, see ya!*
Right!