What is the Semantic Web?
Simply put, the Semantic Web is a more intelligent version of the web as we know it. Instead of just focusing on the structure of online data, it is more concerned with the semantics (i.e., meaning) of such data and how they relate to one another, as provided by descriptions and context clues that a machine such as a web crawler can read.
It is built on three primary technical standards:
- Resource Description Framework (RDF) – This is the data modeling language used by the Semantic Web. Whatever data it works with is represented and stored in this format.
- SPARQL Protocol and RDF Query Language (SPARQL) – As the name implies, this is the official query language the Semantic Web uses. It is designed to fetch data from across different systems.
- Web Ontology Language (OWL) – This is the knowledge representation (KR) or schema language used by the Semantic Web. It allows you to define fully composable concepts (i.e., can be assembled and combined with other concepts in virtually unlimited ways) for whatever purpose.
The Semantic Web is the brainchild of World Wide Web inventor Tim Berners-Lee. It was the result of the realization that while the search engines of old are able to index web pages, they cannot do much in terms of identifying and retrieving exactly what the user is looking for. This is why, at the time, websites can just write keywords repeatedly all over their pages to rank high on search engine results pages (SERPs).
How exactly does the Semantic Web impact online search?
Most of the traditional web makes use of unstructured data, which are basically bits and pieces of data that are not classified or arranged in any sort of meaningful or predefined way. The result is that while search engines could pick out patterns within web pages using the keywords available, they had no way of interpreting them.
In contrast the Semantic Web is designed to significantly improve the quality of search results because, again, it relies on machine-readable descriptions, context clues and logical methods of storing and linking data to give keywords meanings and determine just how relevant they actually are to the user.
The concepts behind the Semantic Web also have applications outside of search engines. In fact, they are currently being used in online advertising, analytics and database management. But that’s not all you can do with them. In theory, they can also be applied to pretty much any form of technology that uses linked data. For the purposes of this post, however, we’ll focus solely on how they are used for semantic search.
Semantic search and triplestores
Semantic search, as the name implies, is just online search in the context of the Semantic Web. It primarily uses what are called triplestores to import and export structured data and ultimately give meaning to the search keywords people use.
A triplestore, on the other hand, is a database of triples, which are essentially just collections of linked entities (we’ll talk about entities in greater detail in the next section) that typically have all the components of a basic sentence: a subject, a predicate and an object. It is typically presented in the form of Uniform Resource Identifiers (URIs).
Take this sentence, for example: “Gretsch makes guitars.” In this case, “Gretsch” is the subject, “guitars” is the object and “makes” is the predicate that links them together. For triples, the subject and the object are both considered entities and the predicate is the relationship between the two.
What is an entity and what is its significance to semantic search?
Entities are basically just a person, place, object or pretty much any kind of concept that we can assign some sort of meaning to. In terms of semantic search, entities, as we have seen in the preceding section, do not exist alone but are given context based on how it relates to other entities.
Say you searched for the band Hall and Oates. Instead of just looking up web pages that contain the keywords, a semantic search engine would try to figure out what exactly these words mean based on all the other entities associated with them like song and album titles, music videos and tour dates. By doing so, it can then pick out exactly the content you need.
Enter: Google’s Knowledge Graph
Of course, we cannot talk about search without mentioning Google. The world’s leading search engine has also fortified its semantic search capabilities by developing its very own proprietary knowledge base called Knowledge Graph. It basically collects all the information it can get – from various credible sources online – on the keywords people use and then figures out how they are connected to each other.
Simply put, this knowledge base is designed to help Google’s search engine determine just how relevant a web page is to the user based on what it knows about the keyword they used, the website which the page is a part of and the actual contents of the page. In effect, it has forever changed how the company’s web crawlers interpret keywords. Instead of seeing a specific search term as no more than a meaningless set of characters, they now view it as an entity that has a clear context and meaning.
What does all this mean for the SEO professional?
Well, more and more websites are becoming more Semantic Web-friendly. Instead of just focusing on which keywords to use, content creators and SEO professionals are now crafting, storing and linking content in ways that are more meaningful to web crawlers. This means that if you stick to traditional SEO techniques, then sooner or later, you’ll inevitably end up losing the SEO game.
But traditional SEO techniques are still far from being obsolete
Keep in mind that no matter how great your SEO and semantic web optimization (SWO) strategy is, the thing the user gets to see first is still the link to your content. This means that while it is important to make sure that your content is machine-readable, you still need to make sure that it is as attractive as possible when it appears on SERPs.
Take a look at the two screenshots below:
Screenshot 1: Example of a basic search result.
Screenshot 2: Example of a more detailed search result.
Google pulled both of these up when we searched for “brownie recipe.” Which one will you be more likely to click? Our guess is that unless you are actually looking for the most popular brownie recipe on Pinterest, you’ll most likely choose the second one. And why wouldn’t you? It captures your attention by showing you a photo of super fudgy brownies and then reels you in by giving you a preview of the actual recipe that piques your interest but leaves you hanging and eager to learn more.
What do you need to do?
The first thing you should do is craft your content to be more Semantic Web-friendly. This means making it crystal clear to web crawlers what exactly it is you are talking about. For instance, if you are running a website about “magic” (i.e. magic tricks), you need to make sure that the way you write your content makes it obvious that you are not talking about, let’s say, the deck-building game, the song by hip-hop artist B.o.B, the NBA team or the band.
A great way to do so is to focus on writing for people instead of search engines because it forces you to organically use keywords that are 100% related to the topic that you are talking about instead of just randomly inserting keywords in the hopes of getting Google’s attention.
But, if you want to be a bit more nitpicky about the whole thing, you can also use a service like texttools to not only figure out exactly what keywords are most associated with the topic you are writing about, but also how much of each keyword you should use for best results.
The second step is to use schemas to mark up your pages. A schema is a type of vocabulary that describes microdata. Microdata, on the other hand, is a type of specification designed to nest metadata (i.e, data or information that describes other pieces of data) within the content of a web page. It allows web crawlers, search engines and browsers to make sense of what a piece of content is all about by making it not only human-readable, but machine-readable as well.
Schema.org, an open community collaboration between Google, Yahoo!, Bing and Yandex, has schemas that do not only make sense to their respective search engines, but are also usable in other channels such as email and even Pinterest. This ultimately allows webmasters and SEO professionals to optimize their content for a wide variety of platforms in one go – which is exactly why you should be using it.
As of April 2017, Schema.org’s core vocabulary is made up of a total of 114 Enumeration values, 589 Types and a whopping 860 Properties. Needless to say, we won’t be able to list all of them down in this post. But here’s an example to give you an idea of how it works:
Basic HTML5 markup:
TR Science and Technology Hub. My friends call me Ty. You can learn more about me
head of the New Technologies Department
TR Science and Technology Hub.
My friends call me
You can learn more about me at
As you can see, the second block of code uses Schema.org vocabularies to tag and describe all the important details in the section and make them machine-readable, all to make them more semantic search-friendly.
If you wish to learn more about schemas and Schema.org, you can visit the official website of the project by clicking here.
Now, once you get this whole thing down, you can move on to making sure that your content is as attractive as it can be once it comes up on SERPs. A great way to do so is by using rich snippets.
What are rich snippets?
Well, before we can answer this question, we must first define what snippets are. Basically, snippets are the descriptions you see under search results, as in the example below:
Rich snippets are practically the same thing but they show more information and fall into nine distinct categories:
- People – provides the searcher with a quick overview of useful information regarding a person, such as their profession, biography, family and notable accomplishments
- Products – displays relevant information about a particular consumer good
- Reviews – provides the searcher glimpse of what other people think about a certain product, service or establishment
- Events – displays a summary of important event details like schedules, venues, titles and performers’ names
- Recipes – gives the searcher a quick preview of the actual recipe along with a photo of the finished product
- Local Businesses and Organizations – provides the searcher a quick overview of what a business establishment has to offer
- Music – displays relevant album information like song lists, release dates, genres, nominations, awards won, all in addition to each album’s cover
- Videos – shows a thumbnail of the video along with a quick description and other relevant details
- Software – works essentially the same way as the products rich snippet but exclusively for application and software products
Again, as we have demonstrated earlier, searchers no longer just click the top items on SERPs. Instead, they are now choosing those that they deem most relevant based on the available information under each link. Rich snippets allow you to capitalize on this in a big way, so be sure to use them whenever you can.
The bottom line: the only way to conquer the Semantic Web is to combine SEO and SWO techniques
You can no longer rely solely on traditional SEO techniques if you want to stay ahead of the competition in the Semantic Web. You need to also start marking up your pages to support semantic search. Keep in mind that semantic search engines need all the help they can get when it comes to understanding your content. Otherwise, they’ll look elsewhere for answers.
Think of the whole thing as baking a cake where SWO is the cake and SEO is the icing. You wouldn’t serve either component alone, right? You need both to get the job done, but you always start with the cake (i.e., great Semantic Web-friendly content) and then cover it up with the icing (a solid SEO strategy).