Knowledge graphs vs Relational databases

Knowledge graphs are known to allow users to store data in a more flexible manner and allow users to interact with tangled data more quickly in comparison to relational databases. Relational databases are a more established form of storing data, by providing pre-defined relationships that data can fit into. In this article, we explain how to set up a knowledge graph on top of your relational database and how it can provide you with deeper insights.

‍

What is a knowledge graph?

Knowledge graphs allow data to be stored more flexibly as data can be stored as nodes and triple stores. Nodes allow multiple relationships to be built upon them and triplestores enable these relationships to be built, they can be used to describe pretty much any kind of relationship which is why they are flexible. Legislate have written a more in depth article about knowledge graphs, but one of the primary advantages of a knowledge graph is that we don’t need to fit our data into predefined relationships. We can define new relationships for this data if a suitable relationship doesn’t already exist, and it is simple to build between new and existing data, without having to change the schema or having to think about the effects that new data could have on existing data.

‍

What is a relational database?

A relational database is a collection of data stored in a database, with predefined relationships between individual pieces of data. Since the relationships are predefined, the data has to “fit into a box” and under a heading or description that has previously been defined. Each row has a unique identifier so that the columns that follow store properties for this specific data point.

‍

‍

This structure allows data to be categorised and it’s easy to use. Many companies use relational databases because of their simplicity and security, and because of this reason SQL is used pretty much everywhere in order to interact with data and is essential for data analysis. The disadvantages of relational databases however, are that their structure limits the way in which data can be arranged and can cause problems when changes need to be made to a dataset since everything is arranged in a particular manner. Although they allow you to store large amounts of data and numerous tables to be created, as the number of tables increases, so does the time taken for a query to be executed, so relational databases can also be slow when querying.

‍

Can relational databases and knowledge graphs be used together?

They definitely can be. Knowledge graphs focus on data that is connected whilst relational databases focus on storing data without deep analysis. An easy example of using them together would be to use a relational database to store key information whilst using a knowledge graph for analysis.

‍

For example, in the context of a dating app, we could store all the personal information of the users in a relational database, such as their names, date of birth and location. But we could store their hobbies and characteristics they value in a knowledge graph so that we can identify possible connections between people. This could be done by using rule-based reasoning to suggest that if two people have more than two or three common interests we could suggest they go on a date with each other. However, it’s still important that we have the relational database in order to store these individual people and their individual data points as key values that have a unique identifier in the relational database.

‍

From the example above, we could suggest that Mandy Collins and Jack Smith appear to each other as a match, as they have similar interests (entertainment and sports) and similar values (family oriented).

‍

Clearly, knowledge graphs and relational databases have their own sets of advantages and disadvantages. Knowledge graphs are useful if your data model is constantly evolving, because you can easily add in relationships or build upon the relationships that already exist, by using more triples or rule-based reasoning. But if you aren’t too interested in data analysis and prefer key value storage, relational databases might be the easiest option to implement for you. Nonetheless, both options can definitely be used together to support data modelling and gain more of an insight into your data. TextMine is doing exactly this to provide users with a deeper insight into their contracts with a searchable knowledge base, and to help build smarter contracts.

‍

About Legislate

Legislate allows users to easily create, sign and manage contracts that are prudent and fair. Legislate’s platform is built on its patented knowledge graph which streamlines the contracting process and aggregates contract statistics to quickly unlock valuable insights. Legislate’s team marries technical and legal expertise to create a painless, smart contracting experience for its users. Book a demo or sign up today to find out more.

The opinions on this page are for general information purposes only and do not constitute legal advice on which you should rely.