What is spend data classification?

My background is specifically in spend data classification and working with procurement teams, so this is where my knowledge and experience comes from. Whatever area of data you work in, this chapter will give you information and tips to help you work more efficiently and accurately.

Turning to trusty Google to get a definition for spend data classification, the first thing that came up was my very own definition! My definition states that spend data classification is the categorisation of financial data into groups or ‘buckets’ of spend using a taxonomy or category tree.

The next best definition I found was from the cloud-based spend analytics solution Accelerated Insight (2021), which states ‘Spend Data Classification is the process of grouping spend data for similar goods or services, assigning them to predefined categories using a Taxonomy document’.

If I broaden the search to data classification, which covers a much wider range of different types of classification, Technopedia provides a good definition:

My definition of data classification would be similar to my definition of spend data: data classification is the categorisation of data into groups or ‘buckets’ using a taxonomy. I want to define spend data classification in business terms, not technical terms. The biggest area of error in data comes from people inputting data/information who are not experts in the field and are intimidated by the technical terms used when working with data.

Spend data classification requires a different skill and knowledge base than master data classification, because classifying spend data is subjective and very much depends on the context. Rarely is there a black or white, yes or no answer in spend data classification. There can be multiple correct answers for the same supplier and it's all about agreeing to those standards internally, especially if you have multiple people working on the data.

We all like a good horror story, especially a data one. It's like when you pass an accident on the road – you can't help but slow down to have a look and then you’re thankful it wasn't you. Data horror stories are just like that. It gives us comfort to know that we are not the only ones suffering and that there might be people in even worse situations out there.

This chapter will be a little bit different. Instead of just slowing down to look, I’m going to stop and help. It's inevitable that these scenarios will happen at some point so I’m going to suggest what you might be able to do if that situation occurs. You have to finish the book on a happy ending, right?

Let's start with a couple of stories that made the news.

Scenario: Edinburgh children's hospital

There was a new £150 million children's hospital in Edinburgh, Scotland, that had to delay opening in 2019 due to a spreadsheet error from 2012. Not only did it halt the opening of the hospital, but it resulted in £16 million worth of remedial action to correct the error.

So, what happened? Well, according to BBC News, the Grant Thornton report stated:

When the project went out to tender in 2013, one of the bidding companies spotted the error and corrected the spreadsheet when they submitted their bid. They did not win the bid and their correction was not picked up by the team evaluating the bids. In addition to this, there was another error spotted within the same spreadsheet in 2016, but this one was not picked up and an independent tester also failed to pick up on these errors.

In the previous chapters, I’ve shown you how to do everything in Excel. This is because it's a widely used tool and I want to make data classification, normalisation and cleansing as simple and accessible as possible for everyone. There's no denying it, Excel isn't necessarily the easiest option! It's cumbersome, slow and open to a lot of errors, especially if you’re new to using it.

Alternative tools

There are tools other than Excel that you can use, but unfortunately many of them involve tiresome processes and still require editing of the source data separately. For example, in place of using pivot tables to check the quality and accuracy of your data, you could use a visualisation tool such as Power BI, Tableau, Qlik or one of the many others available.

Visualisation tools plug into the data source, i.e. Excel, SQL or Python. Some might help you with parts of the cleansing, but you’ll still have to work in two separate tools to get the job done. This can be frustrating, timeconsuming and it can be hard to trace the source of any problems.

While I try to stay as system agnostic as possible, I have only found one solution so far that gives the ability to view and edit the source data in the same window as the visualisation and update in real-time. There may be other solutions out there that I’m unaware of and I’m always looking for ways to improve my processes so I encourage you to take the methods from this book, improve on them if you can and let me know!

The more options that are available for people to work efficiently with data, the better the quality and accuracy. The most important thing for me is to get as many people working with, cleaning, classifying and normal - ising data as possible, however that may be.

Omniscope

The tool that I have been working with for almost a decade is called Omniscope. From this, I have created my own proprietary methodology for efficiently and accurately normalising, classifying and cleansing data.

The answers that each political community finds to the law reform questions posed by AI may differ, but a near-term threat is that AI systems capable of causing harm will not be confined to one jurisdiction – indeed, it may be impossible to link them to a specific jurisdiction at all. This is not a new problem in cybersecurity, but different national approaches to regulation will pose barriers to effective regulation exacerbated by the speed, autonomy, and opacity of AI systems. For that reason, some measure of collective action is needed. Lessons may be learned from efforts to regulate the global commons as well as moves to outlaw certain products (weapons and drugs, for example) and activities (such as slavery and child sex tourism). The argument advanced here is that regulation, in the sense of public control, requires active involvement of states. To co-ordinate those activities and enforce global ‘red lines’, this chapter posits a hypothetical International Artificial Intelligence Agency (IAIA), modelled on the agency created after the Second World War to promote peaceful uses of nuclear energy, while deterring or containing its weaponization and other harmful effects.

The increasing autonomy of AI systems is exposing gaps in regulatory regimes that assume the centrality of human actors. Yet surprisingly little attention is given to what is meant by ‘autonomy’ and its relationship to those gaps. Driverless vehicles and autonomous weapon systems are the most widely studied examples, but related issues arise in algorithms that allocate resources or determine eligibility for programmes in the private or public sector. This chapter develops a novel typology that distinguishes three lenses through which to view the regulatory issues raised by autonomy: the practical difficulties of managing risk associated with new technologies, the morality of certain functions being undertaken by machines at all, and the legitimacy gap that is created when public authorities delegate their powers to algorithms.

The rule of law is the epitome of anthropocentrism: humans are the primary subject and object of norms that are created, interpreted, and enforced by humans – made manifest in government of the people, by the people, for the people. Though legal constructs such as corporations may have rights and obligations, these in turn are traceable back to human agency in their acts of creation, their daily conduct overseen to varying degrees by human agents. Even international law, which governs relations among states, begins its foundational text with the words ‘We the peoples…’. The emergence of fast, autonomous, and opaque AI systems forces us to question this assumption of our own centrality, though it is not yet time to relinquish it.

As AI systems operate with greater autonomy, the idea that they might themselves be held responsible has gained credence. On its face, the idea of giving those systems a form of independent legal personality may seem attractive. Yet this chapter argues that this is both too simple and too complex. It is simplistic in that it lumps a wide range of technologies together in a single, ill-suited legal category; it is overly complex in that it implicitly or explicitly embraces the anthropomorphic fallacy that AI systems will eventually assume full legal personality in the manner of the ‘robot consciousness’ arguments mentioned earlier in the book. Though the emergence of general AI is a conceivable future scenario – and one worth taking precautions against – it is not a sound basis for regulation today.

This chapter turns to the possibility that the AI systems challenging the legal order may also offer at least part of the solution. Here China, which has among the least developed rules to regulate conduct by AI systems, is at the forefront of using that same technology in the courtroom. This is a double-edged sword, however, as its use implies a view of law that is instrumental, with parties to proceedings treated as means rather than ends. That, in turn, raises fundamental questions about the nature of law and authority: at base, whether law is reducible to code that can optimize the human condition, or if it must remain a site of contestation, of politics, and inextricably linked to institutions that are themselves accountable to a public. For many of the questions raised, the rational answer will be sufficient; but for others, what the answer is may be less important than how and why it was reached, and whom an affected population can hold to account for its consequences.