https://data.blog.gov.uk/2016/12/05/how-we-describe-data-infrastructure-products-and-services/

How we describe Data Infrastructure products and services

The Data Infrastructure team has done a lot of work lately to improve the way we talk about our products and services.

One of the ways we’ve done this is by defining the terms we use to describe the things we’re building. Previously, a lot of the teams within Data Infrastructure used similar words to describe different features of different products. There was a risk this could cause confusion, so we decided to come up with clear definitions of commonly used words and make sure we use them consistently.  

Since agreeing a common language for registers and clarifying how we talk about addresses, we’ve been looking at how we describe the work we do around publishing, managing and sharing data. This involves collecting, organising and defining more words, and compiling them in a list.

While doing this, we realised that a lot of the words we use aren’t necessarily exclusive to the products and services we build. People who work with data in other departments and agencies are likely to use these words day-to-day too.

We wanted to know how people across government use and define these words. For example, does “datatype” mean one thing to us and something completely different to someone else? To find out, we’ve decided to make our list available for everyone to see. You can now find it on Hackpad.  

We’d love to get your feedback on this so we can continue to improve it. Please leave your comments below, or on the Hackpad itself.

1 comment

  1. Ben

    I like your list. One part that I find interesting (and am not entirely sure I agree with) is the differentiation between entry, item and record.

    My understanding of the term entry when viewing a register would be that each row in a registry in table format would be an entry. Your definition indicates that an update of that row would constitute a separate entry. I'm not sure I agree. My perception is that an entry would be more similar to what you have listed as a record.

    However, the term record is problematic when you involve records management. People could become confused when you refer to a record as a line in a registry, and then also refer to the document containing the registry as a record. But if you consider that an entry changes over time, what you are recording is the changes in those entries. Therefore, it makes sense to me that an iteration of an entry within a registry is a record. That way a registry at any one point in time would be a record of the entries as they stood at that point.

    For the term item, it seems to me means referent, which is not actually contained in the registry. It's not made up of the fields and values. To me, the entry consists of the fields and values that describe the item. The item is external to the registry. So, for example, your country registry contains information about countries. An entry for Great Britain describes the item Great Britain. However, you don't/can't actually fit the country of Great Britain into a registry. But what you can do is add an entry in the registry to describe it.

    Link to this comment