Recommender System with Open Standards

The following data story demonstrates one of the solutions developed and presented at the NEN Hackathon 2022 with the aim of improving the findability and searchability of NEN norms and standards. The following are live SPARQL queries demonstrating the types of search functionalities that are now possible having transformed metadata about NEN norms to linked data. To see the data on which these queries are being performed, please visit the dataset page.

Functionality 1: "Related to this topic"

When we search in an online store, for example, bol.com or amazon.com, for "Harry Potter":

We automatically receive suggestions for other products that we may be interested in as well:

As part of the NEN hackathon, the Kadaster team sought to replicate this functionality on NEN norms and standards to improve the ability of users to find norms and standards associated with a specific topic or based on previous searches. The following query demonstrates this functionality by providing insight into what norms and standards are associated with each other and, therefore, what may be 'recommended' for a user by NEN based on a users search activity. The following query also provides additional insights such as how many common words two norms share and how old a recommended norm is. This supports the user in accessing relevant norms for relevant situations and in making informed decisions when implementing norms.

Figure 1: Query demonstrating the results of a recommender system implementation