Featured field and capability
Linked Data
Model-Based Systems Engineering
Featured industry
Services
Featured case
Data Libraries
Managing, integrating, and sharing vast amounts of data is crucial for organizations in different industries. But how can information from diverse sources—such as project specifications, asset data, or patient records—be connected and made machine-readable?
The answer lies in Linked Data, a set of principles for connecting data points using web standards. It’s based on a universal data structure, allowing data to be accessed and related as one coherent web. Using Linked Data successfully requires knowledge of modeling languages like RDF(S), OWL, and SKOS, along with query and validation tools like SPARQL and SHACL.
In this blog, we’ll explain key concepts—such as Linked Data, RDF, SPARQL, SHACL, RDFS, OWL, SKOS, and URIs—and show how these technologies work together and can help businesses improve data management.
Linked Data is a set of standards that allow for structuring and connecting data across the web, ensuring that different systems can understand and use this data. For example, a hospital can link patient data with medical records, or an engineering firm can connect project specifications with building information models (BIM).
This concept is part of the semantic web, where data is linked in a way that machines can also understand and interpret without the interference of humans, making it easier to share and analyze. Linked Data relies on a standardized format called RDF to ensure everything fits together.
So, you can use Linked data when you want to organize data, share information, or connect multiple data sources across different systems without manual intervention. It helps to collaborate better between stakeholders or smarter and broader insights. When using Linked Data, everyone can access the correct information at the right time.
Image 1: Connected web of information
RDF is a standard model and the format of Linked Data. Think of it like grammar rules for sentences. It structures data into simple statements called ‘’triples’’, consisting of a subject, predicate, and object. Imagine trying to describe a piece of information like “Building X is located in City Y.” RDF organizes that into three parts: Building X (subject), is located in (predicate), City Y (object). Since it is a standard format, data from various sources can be linked and understood consistently.
RDF is used when you need a clear and consistent format to describe relationships between pieces of data. This could, for example, mean linking project phases with specific activities. It helps you standardize data, and have better communication between systems. This might mean less time wasted on data entry or ensuring that different systems can share data more easily because it has the same structure.
Image 2: Format of subject predicate object
RDFS is a data-describing framework that extends RDF. It gives data additional meaning by specifying relationships between different types of data. RDF is the way to store and share information on the web, and RDFS helps by adding more rules for how to describe this data.
RDFS tells you what types of structures (or “classes”) exist, like “Bridge” or “Engineer,” and how they relate to each other, like saying an “Engineer” designs the “Bridge.” It uses something called “domain” and “range” to show these relationships. For instance, if you have a property like “is designed by,” RDFS would say the domain is “Bridge” (meaning the designer is connected to a bridge) and the range is “Engineer” (meaning the designer must be an engineer).
Unlike some other systems, where you first define structures (like a “Bridge”) and then give them properties (like “designer”), RDFS focuses on defining the properties first and then connecting them to structures. This makes it easier to extend the description of existing resources.You can use RDFS when you want to create more structured, hierarchical data. It can, for example, help categorize different types of materials and project phases or define relationships between solutions and providers. RDFS helps businesses structure their data in a way that makes it easier to analyze and manage. This means better organization and categorization of data.
Image 3: RDF Schema
OWL helps define complex information about things and their relationships in a way computers can understand. It allows computer programs to check if the information is consistent and discover new facts from the data. OWL documents, called “ontologies,” can be shared on the web and connected to other ‘’ontologies’’, making it easy to link and reuse knowledge.
For example, imagine you’re working on a project to build a bridge. In OWL, you could create an ontology that describes the different parts of the bridge—such as “support beams,” “cables,” and “deck.” You could also define relationships, like how “support beams” must be made of steel and how the “deck” must be supported by both “beams” and “cables.” Using OWL, you can also say that an “Engineer” is responsible for overseeing the project.
OWL allows computers to check these relationships. For instance, if someone tries to use a material other than steel for the “support beams,” the computer can flag it as a mistake. OWL makes sure that all the information fits together logically, helping you make better decisions and catch errors early.
To sum up, OWL is used when you need to create complex, detailed relationships between different types of data. This could mean mapping out relationships between materials, suppliers, and projects. It helps you and your computer to organize and understand complex information, enabling more informed decisions.
Image 4: OWL
SKOS is a standard data model for sharing share and linking knowledge via the web. It’s designed to make it easy for people to link different types of knowledge systems, like taxonomies (which classify things into categories) and thesauri (which list words and their meanings). These systems often have a similar structure, and SKOS helps bring out that similarity so that they can work together. The advantage of SKOS versus OWL is that it doesn’t generate data as much as OWL will do. For generating data, the input data needs to be of a very high quality. If that is not (yet) the case, SKOS can help you get there.
SKOS gives a simple way to make existing knowledge systems part of the semantic web. It also provides an easy-to-use language for creating new systems to organize information. You can use SKOS or combine it with more advanced tools like OWL to describe even more complex knowledge.
In SKOS, “concepts” are given unique names ‘’URIs’’, labeled with words in one or more languages, and can be grouped into categories or linked to other concepts. These concepts can be put into hierarchies (like “animals” →, “mammals” →, “dogs”) or networks of related ideas, making it easier to share and connect knowledge across different systems.
SKOS is used when you need to create a controlled vocabulary, ensuring consistency in how data is labeled and categorized. Because of the consistency across your data, it is becomes easier to search, retrieve, and analyze.
Image 5: SKOS
SHACL is a language that you can use to validate if data in an ‘RDF graph’ follows specific rules or conditions. These rules, called “shapes,” are also stored in RDF graphs. It is very similar to OWL, as both describe what relations an element should have. However, where OWL is more beneficial for making your data collaborative, SHACL shines in validating in-house data. Want to learn more?
The SHACL validator needs 2 kind of graphs: shapes graphs, which contain the rules, and data graphs, which hold the information that needs to be checked. SHACL compares the data graph to the shapes graph to ensure the information meets the conditions. Beyond just checking data, SHACL can also be used for building user interfaces, generating code, and integrating data from different sources.
You can use SHACL to validate whether your data is correct and meets the required standards. It helps organizations to maintain high data quality, reducing errors and ensuring that information is reliable and fit for purpose. This can for example lead to fewer project delays caused by data issues. Finally, you probably want a combination of all of the vocabularies above to describe your data. Each of these vocabularies have their specific purpose, but the sum is bigger than its parts.
Image 6: SHACL shapes
SPARQL is the query language to retrieve and manipulate data stored in RDF format. You can use SPARQL to ask very specific questions about your data, like “Which buildings in Project X use material Y?” or “What width should this road have according to specification X?”
SPARQL is used when you need to perform advanced searches across large linked datasets or multiple datasets together. It allows you to pull data from different sources as a coherent whole with only one query. This helps you to get quick answers from your data, saving time and giving a complete overview.
Image 7: Example of a SPARQL query
An URI is a unique address that helps identify pieces of data, whether they are real, like a person or place, or something online, like a webpage. Think of it as a unique code that tells computers exactly what something is. things on the Web, like people, places, or concepts, using systems like RDF. RDF uses URIs to describe things and their relationships, making it easy to share and connect data.
One type of URI that you probably know is a URL, like a web address. A URL not only identifies something but also tells you how to find it, like “www.example.com” for a website. URIs are used in many web technologies (such as RDF), and are essential for organizing and connecting data across the Web.
URIs are used everytime you create Linked Data. They ensure that each piece of information is uniquely identifiable and easily retrievable. Thereby, it becomes easy to locate specific data and retrieve them quickly, improving operational efficiency.
Image 8: URI
Turtle (TTL) is a way to write and store data that computers can understand and is used with RDF. Other formats for storing RDF data include N-Triples, JSON-LD, and RDF/XML.
RDF uses “triples” to represent information. As earlier explained, a triple has three parts: a subject (what you’re talking about), a predicate (the relationship), and an object (the thing being related to). Each of these parts is identified by a unique web address (URI). TTL makes it easy to create these triples and even allows you to shorten them by using more straightforward names.
For example, if you’re storing information about a movie like Star Wars, you could use a Turtle triple to describe the relationship between the movie and its main character.
TTL is used when you want to store and share data in RDF format in a simple and human-readable way. It’s helpful when you need to represent triples (subject, predicate, object) and want to work with Linked Data or when you need to manage large sets of RDF information more efficiently.
Image 9: TTL file
JSON-LD (JavaScript Object Notation for Linked Data) method of encoding Linked Data using a format called JSON, which is often used to store and share data. It is similar to the Turtle format above but is more comparable to a data format developers already know (at the cost of being slightly more verbose). The goal of JSON-LD is to make it easy for developers to take regular JSON data and turn it into Linked Data without much extra work. JSON-LD looks very similar to regular JSON, so it’s familiar to people who already use it.
JSON-LD is used when you want to organize and share Linked Data in an easy-to-work-with format, especially if you’re already using JSON. It’s great for using to turn regular data into Linked Data without much hassle and is often used in web applications and APIs.
So, to sum up, you can use JSON-LD when you need a lightweight, familiar format to connect your data on the web. It’s beneficial for developers who want to integrate Linked Data into existing systems with minimal changes while keeping it easy to read and understand.
Image 10: JSON-LD file
Linked Data transforms the way businesses manage and connect their information. By organizing complex data, breaking down silos between formats, and boosting collaboration across systems and teams, Linked Data makes it easier and more efficient to integrate, explore, and connect data sets.
Want to learn more about Linked Data topics? Explore our latest Linked Data case study and discover how to manage and optimize your data.
