July 16, 2015 Leave a comment
Today the web delivers information to us in the form of web pages – mainly HTML documents that are linked to one another through hyperlinks. It’s by scanning and following hyperlink after hyperlink that Google indexes and maps the entire world-wide web.
The sweet thing with HTML documents is that both humans and machines can read them. Whereas humans can read the content as such, interpret it and consequently link meaning to the words on the page, machines need to look at some clues to judge the content of a web page. Clues that SEO specialists truly understand. So they put the keywords in places where machines go looking for clues.
But as the web evolves, technologies alter. And moving beyond HTML to communicate information to machines might have a tremendous impact.
HTML is still important – and it will always be, most probably – but the importance for machine interpretation is in decline thanks to semantic technology.
The web has promised semantic for a while now, but nowadays it seems to hold true to the promise. For SEO, it means keywords are no longer the key element of SEO – as proven by Google’s default ‘not provided’ keyword parameter. The new normal for SEO is to do Semantic Web. It means marking up your content in a much more clever way then we previously did with a keyword-driven html-documents-based approach.
Semantics turns Google into both your start and endpoint of every search journey.
As happens with most searches these days, let’s start at google.com to explain what it is about. And as happens with most searches on google.com these days, this is exactly where your search journey ends as well. Why? Because Google provides you the things you need on its own page, not by redirecting you to a scanned page somewhere on the internet.
How do you think Google Shopping allows for online buying on Google at a retailer’s store? Or what about booking hotels and flights directly in Google? And what about the nice preview you get on your mobile phone to confirm your flight and hotel booking? Or something as simple as having your location and opening hours in Google?
Yup, Semantics. So, Hello, new SEO! What a bless that people can find you on Google and e.g. buy your service/product directly through Google, see your opening hours and location, …
Let’s take a look at two big (online) players, Lufthansa and Ebookers.com who are already doing semantic web and giving customers a superb experience on Google platforms.
How nice are those emails? What a brilliant customer experience is this? Quite sure you want to repeat this because it makes you feel good.
Many applications from Google, Microsoft, Pinterest and others are already using these vocabularies to power rich, extensible experiences. It’s important for brands and corporations to offer a smooth experience from your company on the Google, Facebook and other dominant web platforms today. It means that you need to get into the semantics game.
Particularly for the “Google Discoverability Purpose”, the semantic tech at work here is Schema.org. Schema.org seems one of the most important RDF schemes for businesses today to understand and deploy. I even would like to call it the new SEO – if only it was to build awareness by provoking.
Schema.org, the most important Semantic Web tech today.
Schema.org is a community that creates, maintains and promotes schemas for structured data on the Internet, on web pages, in email messages and beyond.
The vocabulary can be used with many different encodings (RDF, microdata, JSON). Many applications from Google, Microsoft, Pinterest and others are already using these vocabularies to power rich, extensible experiences. It’s important to provide a smooth experience from your company on the Google, Facebook and other dominant web platforms otherwise you’re lost. It means that you need to get into the semantics game.
But where do semantics come from? What’s the story behind it?
Be aware: here is where I try to explain the tech behind my marketing bla bla above.
The story behind schema.org: Semantic Web for machine readability.
In its very essence, semantic web, marks a shift in thinking from publishing data in human readable HTML documents to machine readable documents. The Web contains lots of information. But it’s hardly constructed from raw data. It’s marked up in HTML documents. The semantic web basically changes this core architecture of the Web.
The semantic web is a way that allows to describe models of data that can further be treated as if everything was in one database. Think of the web as one big jar of data. Think of a web page as a visualization of well-selected data from that jar. Mashup Gallore! You’ve seen it. And you’ll see it more in the future. Why does this happen now?
Why the time for Semantic Web is now.
Regardless of the fact that Semantic Web has been a topic for years now and that it did not get much traction outside the academic realms yet, the time for semantics is now:
- artificial intelligence is coming strong and if we want to make the web work for us, we need a “language” for machines to understand. That language seems semantics.
- internet of things is generating data and if we want machines to communicate with one another (M2M communication) and with humans, we need a language that both can understand. Hello, Semantics.
- the growing usage of marketing automation at multiple digital touch points, makes an intelligent markup language crucial for machines to take over human activities.
- the benefits of automated research of all data humanity has to offer on the internet…
Semantic Web: the tech side, the machine aspect
So far we did not clarify a lot. We agree. We did not do anything else than saying that the web as we know it – html pages linked via hyperlinks – is about to be replaced with a more clever way of organizing, structuring, retrieving and visualizing data and content.
Semantics is mostly defined as a “3.0 web technology” – a method of linking data between systems or entities that allows for rich, self-describing interrelations of data available across the globe.
One might think this is a rather complicated topic. However, it doesn’t have to be. The below learning curve helps you to understand what semantics is about. And how to get there from the ground up.
How to realize Semantics – a roadmap from Graph Data over RDF to Semantics.
Forget everything you know about databases. Because probably you only know hierarchical relational databases. This hierarchical architecture is what the semantic web leaves behind. It starts from a Graph Database instead of a relational and/or hierarchical database.
A graph consists of resources related to other resources, with no single resource having any particular intrinsic importance over another one. In this way, it’s easiest to understand a Graph as a visualisation of a series of statements about how things relate to each other. Let’s clarify things with a data graph example.
A Data Graph example
The below graph (or scheme if you want) actually makes several statements about two objects:
- thing 1 has the name Bengie and has the animal type dog
Bengie is a dog.
thing 2 has the name Bonnie and has the animal type cat.
Bonnie is a cat.
Thing 1 is linked to Thing 2 through a relation “friends with”.
Bengie and Bonnie are friends, despite being a cat and a dog.
Is it only me or are people also constructing meaning in this manner?
Anyway, let’s look at a method do translate this graph into something that can be used in Information technologies. Hello RDF!
RDF: the foundation of semantic web without describing the meaning (semantics) as such.
RDF is a formal manner of describing data graphs so that machines understand the structure behind the graph. An RDF statement always comes with the following 3 dimensions:
The “triple” (subject, object, property) is the essence of RDF, the structural foundation of the semantic web.
In order to thrive in the semantic web, one needs to understand data and content within this non-hierarchical means of data modelling.
Semantic Modeling with metadata from formal vocabularies and ontologies
While the above RDF statements offer a graph-based model for recording and interchanging data globally, it doesn’t provide any clues for the meaning as such of the statements. In other words: there’s nothing semantic in yet. To include meaning, one needs common formats to collaborate. These common formats are realized through vocabulary and ontology:
- vocabulary: terms with a well-defined meaning across contexts
- ontology: defines contextual relations behind a defined vocabulary.
Standard vocabularies, or formal ontologies, are already available for a wide range of subjects: media terms, biomedical terms, scientific terms, etc.
For non-techies it might be the easiest to understand all this as “metadata”; data that describe the data. For SEO’ers it might be easiest to understand that “schema.org is the new meta data in the html head section”.
The real deal: RDFS & OWL, the actual semantic techology
In order to construct meaning in RDF data, one needs to mark them up or annotate them with semantic metadata: RDFS and OWL. And it this point, we’re really talking technical shit. So I think I’ll leave it for now. In fact, I’m quite unable to further explain this. I’m a non-technical guy. But I believed this was interesting. So I researched a little bit. So, yeah, it might be, this post has errors…
The web is undergoing a massive re-architecture. It’s called Semantics.
We’ve pointed out in the introduction that Semantic is getting increasingly important to the world because of the rise of artificial intelligence, internet of things, marketing automation and so on.
But we started from a “Google / Seo” perspective because we believe it’s shows the importance of semantics in the most tangible way. SEO is about getting a good spot in the search enginge result pages for a specific keyword. Clever marketers have already noted that Google’s search engine result page is not longer a gatekeeper. It is not the start of searches. It’s also the end of searches. Making sure your product, brand, service pops up in those end pages. Intelligently mark up your bits and pieces (data, content) with schema.org.