Nick Cavarretta

Internet Marketing Blog Posts

Here are my rantings on internet marketing. I've given some good advice in some articles, and others, well. Take a look for yourself.

The Semantic Web and Linked Data Simplified

If you're one to take part in the world of SEO, or even the internet, you would have heard something called the semantic web at some point. An extension from that would be linked data. If you haven't heard of these terms, then perhaps microformats, or schema. This article is set out to simplify the explanation of the semantic web, linked data and it's role in SEO and development.

What is The Semantic Web?

The semantic web is the extension of the internet that enables computers to understand what we are talking about within data. Computers only understand syntax, which is how you say something. They don’t understand semantics, which is what a sentence means.

If I were to say that “I love SEO”, the semantics means that I really like the topic SEO. If I was to change the word “love” to a heart icon, the syntax changes but the semantics remains the same.

The semantic web is really talking about is organizing and structuring data to enable better communication between machines.

The internet was invented so computers can talk to each other, exchange different document formats: basically to send and receive data. It wasn’t built to understand what we’re talking about.

When the world wide web came along it opened up the worlds documents for sharing. Companies would store their information in a database, where the public could then use the companies search engine to find the information they require. Search engines started using keywords to deliver a search result, and to some extent still do. This is basis of what we learnt “back in the day” when searching, writing and back linking with a specific anchor text.

Image from

Today search engines like Google have implemented their own content based algorithm filters, like Panda and Hummingbird. However, this is not just to fight spam, it’s to help semantics. Now RankBrain looks like Google's promised contribution to the semantic web and AI, promising more relevant search results for users.

Computers don’t understand meaning, only syntax. If we can get computers to learn the semantics of a document, or webpage, then search engines can deliver a much more relevant list of documents from a question or odd search query.

Of course this issue also extends into tone and intention, just like the real world. Check out the Arj Barker joke on how you can't convey tone on the internet. How many times have you received a text, and although you understood the syntax, the intention behind semantics is not quite clear? At least you have the ability to reply for clarification, or if you know the person's character, you'll have a better understanding of the intention or tone behind the sentence.

The web we know today is really about documents where it’s starting to move towards a web of Things. Not only does the semantic web want to know the meaning behind a phrase, but how Things are linked and related to each other.

So the semantic web is really here to help us get what we want. If everything was structured, the semantic web could deliver results with a much higher relevancy, and it won’t have to rely on keywords anymore.

What is linked Data?

When we think about data, we can imagine we’re talking about images, text, video and audio. The web allows us to link from one document to another. Although this is great for humans, robots are still having some confusion in understanding the relationship between links.

The problem computers have with data is they don’t understand it the same way a human would.
So what we do now is use something called microformats to to structure our data, but there is another problem with using that. Nobody can agree on a packaging. So far we can use HTML, JSON, XML, CSV and RDFa.

The other issue is how we link all this data together. There was a platform called Freebase that aimed to collect and store information for Google’s knowledge graph.

Today Wikidata is the biggest player in structured data, taking over Freebase when they closed the doors to submissions.


The best way big platforms are measuring linked data is by using a Graph. Google uses the Knowledge Graph, and Facebook used their Open Graph. Essentially these are the same thing, well, for the same purpose anyway which is linked data.

Let’s look at Facebook Open Graph for a moment. How would we know how Nick is related to Natasha? Well we wouldn’t unless the two have linked their data in the Open Graph.

When Nick added Natasha as sister under the relatives option in Facebook, that told the Open Graph that Nick and Natasha are siblings. This will link Node:Nick to Node:Natasha as Relationship:Siblings. This is all linked data and part of a very large Graph in order for machines to understand the relationship between two links.

If linked data was more advanced, we could go into Facebook and ask a question like “who is Nick’s sister?” where it would reply “Natasha” and display her profile. Open Graph uses RDFa technology for tagging, and encourages people to tag pages or people in data so the computers can understand the link between the information and what the data is about.


Now you have a basic idea on the semantic web and linked data, you can take action in putting everything into practice.

To add relevance to your content and feed the semantic web, I’m going to put emphasis on using words your target audience uses and words your industry influencer uses. Stop thinking about targeting a “keyword” or “phrase” when writing an article, but targeting a specific area of the web.

To do this you’re going to need to build a seed list, which is a list containing all the words you should use in your article. Start this task by opening notepad and coming up with industry specific words. Think about what single words your target audience may use.