The University of Canberra Sport Informatics and Analytics (UCSIA) MOOC I’m working through to augment my dissertation research, invites us in the first module to think about various definitions of informatics and analytics and what distinguishes them. From my perspective I’m also interested in thinking about how Information Science fits in. This blog post covers my thoughts on module 1 of the course and how it relates to my research.
Informatik, Informatics, Information Science?
In the course Keith Lyon traces the origins of of the systemic study of information managing processing from the German term Informatik to the anglicised version Informatics. Michael Fourman, from the Division of Informatics at the University of Edinburgh provides this definition of Informatics (2002):
Informatics is the science of information. It studies the representation, processing, and communication of information in natural and artificial systems.
Now, as someone studying Information Science, that sounds a lot like my discipline so what is the difference, especially as Fourman’s definition of information and processing is quite broad:
Representations include paper, analogue, and digital records of text, sounds and images, as well as, for instance, the information represented in a gene, and the memories of an individual or an organization. Processing includes human reasoning, digital computation, and organizational processes.
The differences between Informatics and Information Science as disciplines are probably increasingly fuzzy and blurred in the age of an expanding infosphere. Even Librarians can’t decide: the Dewy Decimal Classification lists Informatics under both Computer Science (004) and Library and Information Science (020).
That there is a distinction is probably due to slight differences in emphasis derived from their histories and the converging disciplines that formed them. Informatics is probably more closely aligned with computer science, emerging during the 1960s. Information Science’s origins are in the more scientific approaches taken to library management and documentation during the 19th century. So Informatics is 004 and Information Science is 020, but they both are both in roughly the same territory and both examine a terrain that is rapidly transforming.
A quick look at Google Ngram Viewer (date range 1900-2008) shows documentation in use from earlier and remaining predominant with computer science and information science increasing at the same time, around the 1960s, with information science levelling off first. It seems that around the mid-1990s informatics overtook information science. Analytics has seen steady growth with Data Science barely in use still in 2008. Sport Informatics and Sport Analytics are not yet recognised as Ngrams.
Figure 1: A rough look at the history of some relevant terms
The Story of Documents and the Information Communication Chain
At #citylis the history of information is the story of documents. We study how information is instantiated in documents from pre-historical cave art to our current data-rich society, whilst also looking ahead to consider how information technology may be bending the contours of our reality generating new types of immersive documents that are as novel to us as the codex or computer were previously.
Consequently, we adopt a broad definition on information science as a field of study that is necessarily multidisciplinary, recognising that this field overlaps with others. Whilst other fields my touch on the information communication chain (Robinson, 2009), Information Science covers all components in this chain:
- storage; and
Our definition is (Bawden and Robinson, 2012):
Information science can best be understood as a field of study, with human recorded information as its concern, focusing on the components of the communication chain, studied through the perspective of domain analysis
I approach sport as an information domain from within this theoretical context, which might be slightly broader than the common view of informatics and analytics in sport. This was my initial brainstorm of all these terms attempting to fit computation-oriented perspectives with the #citylis world view. This still left me with plenty of questions.
Figure 2: Attempting to make sense of overlapping terminology and disciplines
Daniel Link and Martin Lames have advanced a definition of sport informatics as an interdisciplinary space where computer science and sport science work together (2009, 2009, 2015). In fact they suggest sport informatics is synonymous with “computer science in sport”. Their definition of sport informatics (2015) is:
Sport informatics is a set of multi- and interdisciplinary research programmes which contain parts of sport science and computer science. The subject area is the application of tools, methods and paradigms from computer science on questions of sport science as well as the integration of sport scientific knowledge in computer science.
For me and my research, this is too narrow and neglects the perspectives that Information Science could bring in understanding sport as an information domain. It also is too vague on the range of design and engineering disciplines required to successfully gather data in sport; computer science doesn’t necessarily cover the interface between athlete and technology.
Informatics and Analytics
The UCSIA module also asks us to think about the difference between informatics and analytics. For this I checked on some philosophical definitions of analysis in the Stanford Encyclopedia of Philosophy (Beaney, 2015). The word is derived from the Greek for ‘loosening up’ and it tends to be thought of as breaking something down (decomposition), particularly to understand it better or solve a problem. However in order to do this we may also have to interpret or translate it in order to explain it. If analysis involves breaking down then its converse, synthesis, is often described as building up. Synthesis involves reconstructing the something or problem from the fundamental pieces found during analysis. In Aristotelian Analytics , analysis can be described as a “method for discovery” and synthesis as “a method for proof” (Beaney, 2015),
My favourite definition was provided by Antoine Arnauld and Pierre Nicole:
Now analysis consists primarily in paying attention to what is known in the issue we want to resolve. The entire art is to derive from this examination many truths that can lead us to the knowledge we are seeking.
In my opinion, Informatics is the process by which we find, organise, interpret and share information. This process is at the heart of the domain of meaning and sense making turning tiny bits of data (small pieces) into derived knowledge (big pictures) via information (the arrangement of the pieces). Analysis provides methods for observing and interacting with the small pieces, the structure of things, and Synthesis provides methods for assembling the big pictures, the relationships between things. Analytics could be described as the application of these methods to Informatics.
Sport informatics involves collecting data, things known or assumed to be fact, then arranging it in ways that convey or represent something about sporting performance or participation. Collection can involve the capture of data from sensors or the observation of athletes. Arranging it involves storing the data in an appropriate database and potentially organising it; for example, by cataloguing/coding data or video clips with specific athlete, sport-specific, tactical or outcome metadata.
Interpreting this information, using sport analytics, involves identifying patterns that tells us something about the sport and extends our knowledge of it. It involves both discovering and proving knowledge Tools and techniques that can help with this include data visualisation, predictive modelling and machine learning amongst many others. The successful application of knowledge can be considered wisdom. In sport this might mean improving athlete or team performance, increase sporting participation, deepening fan engagement or developing innovative recruitment or coaching strategies.
To be effective all four levels need to work well together. The right data needs to be collected, arranged so that it conveys the right information, interpreted so the right knowledge can be derived and applied wisely.
My Model of Sport Informatics and Analytics
Taking the UCISA MOOC has enabled me to reflect on my studies with #citylis and compare information science to definitions of informatics that are more common in sport. This has enabled me to reconcile some definitions, at least in my mind, that provides the broad context for my research. I thought about this research and the things that interest me about sport information and combined these thoughts with the various theoretical and definitional influences discussed to create a model of my view of sport informatics that expands on the Link and Lames definition. You could argue that having completed my analysis (broken down sport informatics and analytics into its first principles) this is my synthesis (exploring the relationships between them to assemble a coherent whole).
Figure 3: My general model of sport information and analysis.
In this model my four ‘fields of study’ are computer science, information science, engineering & design and sport science. At the very centre all four intersect as sport informatics and analytics. At various other intersections not an exhaustive pr precise list of overlapping topics, but rather these are some other themes and trends that I am particularly interested in.
For example, the Internet of Things has the potential to turn any physical device or being into a sensor connected to the Internet, capable of sharing its data continuously and being controlled remotely. Where this trend connects with sport you have Athlete 3.0: a connected sports participant whose activity in the biosphere is immediately and continuously projected into the infosphere by a range of sensing devices.
In my model informatics, analytics and data science come at the intersection of computer and information science. When applied to a domain these will always bring an important philosophical dimension and ethical considerations.
Between information science and sport are the theoretical frameworks underpinning my dissertation study of amateur athletes: domain analysis and information behaviour. These are information science methods that allow the informational nature of a particular domain, in the case sport, to be examined.
Whilst also related to computer science, I see virtual reality and augmented reality as predominantly emerging from the engineering and design fields that create the devices and interfaces that cross and blur the boundaries between different conceptions of reality. Within the sport domain these are creating new forms of virtual sport and augmented sport with new options for competing, training and fan engagement. When combined with information science and its interest in documentation this becomes an interested in immersive documents: the new forms these will take and the new literacies they demand.
Up Next: Pattern Recognition and Amateur Informatics
Module 2 of the UCSIA MOOC is on Pattern Recognition and covers what makes data interesting and how we make sense of it. This should take me a bit further out of my comfort zone into the worlds of systematic observation, supervised machine learning and a more explicit look at big pictures and small pieces in sport.
My dissertation research is more focused on the information behaviour of amateur athletes, rather than elite athletes and professionals. I’ll write shortly about how this generalised model of sport informatics fits with the everyday athlete and the information I’m interested in researching in a separate blog post. I’m also working on the design of my athlete survey which is currently being piloted. It should launch by the end of the month so I’d encourage any runners, cyclists, swimmers and triathletes who are interested in taking part in my research to keep any eye out for the launch announcement.