×
Below, we sum up our data resolutions for the year

Data Resolutions

Below, we sum up our data resolutions for the year ahead. Which of these resolutions will you stick to in 2024?

 

Keep Our Records Up to Date

At the Institute of Analytics, our guidelines on data documentation have always been left open to interpretation. On the one hand, there are ethical and responsible reasons why we need to be transparent in what we are doing and transparency requires clear documentation. On the other hand, as an analyst we don’t want you to spend too much time documenting at the expense of carrying out your main role. 

In 2024, it will be easier to produce good records through the use of language generative models. While we don’t advocate putting your proprietary source code into a machine, you can easily get appropriate documentation on generic code from a language generator like OpenAI tools, and then you can add any ground truth data or details in your own words. Don’t forget that GPT can turn bullet points into free text too! 

 

Speak a Common Language

It’s more important than ever that we share our work with our colleagues and involve them in the decision-making process. Not everyone on the IoA team is fully data-literate and we face some of the same cultural upskilling challenges as some of you. 

This last year we received a lot of requests for support with AI reporting and we will be launching our AI Auditing training and resources in the coming year. These will support our data scientists in documenting and our auditors in challenging and critically evaluating those reports. 

In the meantime, we suggest taking a look at our Model Cards to Support Responsible AI on our Website [Link: https://ioaglobal.org/insights/#cheats]. Here we outline the essence of good practice in reporting both the advantages of using data analytics and the risks associated with it. These may be obvious to you and me, but could be unknown for your colleagues. 

 

Stay Up to Date with the Law

It’s not enough to adhere to data privacy laws any more. Although 2023 saw a lot of calls for regulation of AI – and not much governmental response – there are still plenty of laws out there that impact our work and you may not be familiar with all of them such asIntellectual Property Law, International data transfer regulations, GDPR Article 22…

If staying up to date with the law is not an option due to the growing complexity of overlapping regulatory environments and industry standards, then we suggest robust best practice guidelines. Our Governance and Professionalism training is updated regularly to reflect changing understanding of best data practice, and any interim updates are on our Resources page including our Checklist for an Organisational Generative AI Policy [link https://ioaglobal.org/insights/#cheats]

 

Embrace the Beauty of Interoperability

At the IoA, we are in a similar position to many organisations in that we have a number of data sets from disparate sources written to different schemas. We began the process last year of aligning all our data sets to one interoperable format. 

Interoperability is not the most exciting process, but it creates a foundation for innovation to emerge in the future. By aligning to an industry standard and a taxonomy, we know that the work we are doing today will reap benefits in the future with flexible data that we can easily join and gain insight from.