How to Implement the Semantic Web
The World Wide Web is probably one of the greatest innovations we have in the world. It has allowed people to connect and share data is possible with little to no time lag. You may have probably never thought that the www as we know it may need any improvements. Well, it does.
Have you tried to search for someone knowing only their first name? It becomes very difficult.
Lucky, we have a new innovations challenges: the semantic web.
But for those of us not in the information sciences world, we have big questions:
How do we implement the semantic web?
Are there organizations or web pages that are already using it?
How to get started implementing semantic web the easy way
In this article, we will get into these three questions and try to find answers. Let’s dive right in.
We will begin by looking at sites and organizations that have begun to publish and store data using the semantic web concept.
The Czech Republic: Open Data on Government Sites
The Czech Republic has managed to publish data on the trade inspection authority, statistical office, telecommunication and the ministry of finance websites as open data, a core characteristic of the semantic web. Examples of data types published in this format include census data, inspections, bans, and treasury data (financial data, and data from accounting reports).
According to the Czech Trade Inspection Authority, the data from this public body is published via different sets – XLSX, ODS (open document spreadsheets) and CSV. The data is also partially anonymized. No personal information is included. Individual entrepreneur data is presented as aggregated data.
Open data like this is important to the semantic web because it relies on a common data set (or knowledge graph) to help remove ambiguity between terms.
Increase Sales Implementing the Semantic Web
A practical way in which we can apply the semantic web is to help consumers understand products and purchasing patterns. So let’s see how a concept developed by Haklae Kim, Korea Institute of Science and Technology Information can help you understand.
But why would we need the semantic web here?
The simple answer is that the semantic web acts as a sales assistant. Imagine going to buy a night camera on Amazon or eBay to monitor your garden. You need to choose not only between brands but also between products and specifications and features. Unless you have used these cameras for a long time, you may not have a clue of what features to go for. Worse still, you may go for a product with many unnecessary features.
With knowledge graphs, it is possible to break down product features in a way that is easy for consumers to understand.
In the example from Haklae Kim, the product data (features and specifications) can be structured in such a way that is machine-readable. Users can then query this knowledge graph to find information and explanations to inform their purchasing decisions. Possibly even through an automated chatbot.
This simple example is one way we can implement the semantic web on an e-commerce site to make it easier for customers to understand a product’s features and specifications.
You can see this in action in Google’s featured snippets:
So what does an organization need in order to effectively use the semantic web?
A clear communication of change of information. People in the organization need to understand why information management is being done differently.
Set the goals that you want to achieve clearly
Do an analysis of how various stakeholders would be affected, for instance, suppliers
Put together a team to help communicate the vision
Help management understand the benefits that come with using the semantic web
Get the technical team to start (add XML to documents, build ontologies)
Set a staff training schedule and implement
But wait, what about SEO and Semantic Web?
The internet is made of trillions of web pages "stitched" together by links that humans (and search engines) can follow. People don’t normally read HTML code, so we have web design, color, text and UI/UX to help us see what web developers create.
Headings, meta tags and content are then collected by search engines, but they then need to design and create a way to help them gather and interpret that data to say that your website is an e-commerce jewelry store selling, specifically, accessories for men.
This is where structured data comes along and why it’s important for every website to focus on adding structured data to every page. Search engines use graphs where data can be completed and made of entities (subjects) + properties of the entities (property values) + relationships with other entities (arcs), etc.
Schema.org metadata can be packaged in:
Imagine you search for "John Doe". The structured data identifies which John Doe you’re talking about by linking his entity properties to others via relationships such as birthplace, organization, education or job.
Without the help of the relationships defined in the linked data, there’s a lot of ambiguity surrounding which "John Doe" you’re actually looking for.
What Can Semantic Web Do for SEO?
Increase search rank
Increase search traffic
Increase click-throughs by 30% with rich cards.
Building pages is like building relationships between entities. Connecting all information related to a specific yoga session across multiple pages can help strategically stand out on Google search.
Semantic Web Formula: Intent, Interpretation, Knowledge and Answer
You can first create a general organization structured data to help search engines know that you are a organization:
But don’t stop here! Once you’ve added your basic structured data you can check to see if it shows up with any errors using the structured data testing tool in Google Search Console:
If you need to perform an SEO Audit and see if each page of your site contains structured data, you can review them with WooRank.
Adding more value to your site with Schema.org
Reading Schema.org documentation on "Organization" structured data, you’ll be able to find more vocabulary and link data. Adding “Founder” can help search engines as well as people looking for a specific yoga session in New York held by John Doe. Linking Events Schema to your yoga class, can help understand what the event is about, specific time, end date, intended audience, location, etc. Maybe a potential client heard about a yoga session that’s held only on Thursday, by adding PreviousStartDate, you might be able to help your potential client identify and eliminate some doubts regarding if your yoga class is the correct one they were looking for.
Or you might be the only one being sponsored by popular yoga mat, you can also add Sponsor, in your events data.
Wordlift and Semantic Web Implementation
If you’re using WordPress for your site, you use WordLift to simplify your semantic web implementation.
NLP and large knowledge graphs such as DBpedia to suggest general concepts that can enrich your writing.
Creates search friendly 5-star linked data with schema.org that can help improve CTR.
Adds interactive widgets and recommendations to increase dwell time (lower bounce rate by 10%).
Splits content into atoms, improving the navigation on your site.
WordLift analyzes the content of an article and identifies matching entities and topics from large big data graphs such as Dbpedia, Freebase and Geonames or from the user’s internal vocabulary. All extracted concepts are organized in four categories: Who (Person, Organization), Where (Place), When (Event), What (Creative Work, Product, Intangible).
Content is no longer written meaningfully for solely your readers, you can now transform articles into knowledge structured in a linked data graph that machines can process.
As you can see there are sites that are beginning to use the semantic web in publishing their data, though not without any challenges. SEO is not only about adding relevant keywords but helping search engine crawlers collect data, help them link each data and understand them better.
SEO and the semantic web is about providing your website on a silver platter to Google, which then gets you rank higher, increasing both traffic and conversion.