Legal fees can quickly add up, and contracts can be difficult to understand for the unlawyered. Legislate is making contracting easier for the unlawyered, thanks to knowledge graphs. Essentially, knowledge graphs make the manipulation of contract data much easier and speed up the process of contract analysis. This means that instead of having to tediously scan through a contract and identify important clauses or terms, you can use knowledge graphs to do this instead saving you both time and energy. Legislate’s approach also removes the potential of error, because knowledge graphs use logic to assemble clauses together which ensures nothing is missed from a legal perspective. This article walks through the technical process of analysing a contract with knowledge graphs.
How is data stored in a knowledge graph?
Knowledge graphs store information in a rdf triple format. A triple contains a subject, predicate and object. Triples allow you to classify data without providing explicit descriptions, which make them more flexible data storage methods. Data that belongs to the same class can also be grouped and then subsets of data can be created to make it easier to navigate.
Analysing a contract with a knowledge graph representation
Let’s analyse an employment agreement and see how knowledge graphs link the information together. Contracts are stored with a unique ID to differentiate them. For an Employment Agreement, we can classify it as a subclass of a Employment Contract, so as a set of rdf triples, it would look something like this:
And written in rdf:
:legislate_MikeEmploymentAgreement a :EmploymentAgreement;
:EmploymentAgreement :subClassOf :EmploymentContract;
:legislate_MikeEmploymentAgreement a :EmploymentContract.
The contract (:legislate_MikeEmploymentAgreement) will also have other properties attached to it, such as when it was signed, when it was created, when it was completed and who the respective parties of the contract are. Other contract fields might include the job title, line manager, whether or not the role is remote, where the job is located. These values and options will be based on the answers you enter into Legislate.
Deriving knowledge from data
Since Legislate allows you to tailor your contract to your preferences, the terms of the contract will change based on the template you have selected and the values you have entered. This means that the parties’ rights and obligations for a contract will change depending on the information you input. Some standard rights and obligations that would be sorted in the context of an employment agreement would be a holiday period and number of paid holiday days. An employer is required to give their employee a certain number of days of holiday and define the holiday period in which they can take it; in the same respect, employees are restricted by the holiday period in which they can take holidays and are limited to a specific number of days. Although this is a standard requirement of an employment contract, the actual number of days and when the holiday period resets would change depending on the employer and whether they are full or part-time, which is exactly where Legislate and knowledge graphs come in handy. Knowledge graphs would be used to deduce for each individual employee their obligations and restrictions from this information.
This information can then be queried and can also be sorted as an obligation for the employer and a restriction for the employee. The information would be stored according to the type of information it is, in this case “Holiday Period” and “Number of Holiday Days”, hence, if the actual value of the number of holiday days changed from 20 to 25 days, this would easily be updated in the Knowledge Graph and would still be sorted into the according obligations and restrictions classes.
Curious about automated data extraction from documents?
Visualising knowledge graphs
Knowledge graphs can also be visualised, which means you can see the triples on a screen and the relationships between them. Being able to see the visual knowledge graph yourself may help you think of further links or relationships in the data, hence allowing you to further query and analyse the data. It proves how powerful knowledge graphs are in representing data and how deeply they can be explored to reveal new insights or outlooks.
Legislate is an early-stage legal technology start-up which allows large landlords and small businesses to easily create, sign and manage contracts on their own terms. Legislate’s patented knowledge graph approach is unlocking the full potential of contract data and bringing robust contracting to the unlawyered. Legislate’s team marries technical and legal expertise to create a painless, and unique, contracting experience for its users. Legislate is backed by Parkwalk Advisors and Perivoli Innovations and angel investors.