Information visualisation is an essential tool in data-science, but the lack of a theoretical foundation currently prevents visualisation science to make substantial progress and develop solutions for the epistemological challenges posed by Big Data.
Starting from the current state of the art in formal logic and the philosophy of information, the prospects of a new foundation for information visualisation are explored. This should lead to a model of the information-lifecycle in visualisation that sheds light on trade-offs in design decisions, gives a unified account for reasoning and communication with visualisations, and explains why and how information-visualisation allows us to climb the Data-Information-Knowledge hierarchy.
Given the epistemic challenges in science and in policy-decisions, substantial attention is also devoted to what can go wrong with the use of information-visualisation, which requires the development of an account of mis- and disinformation, and of fallacious reasoning based on computer-generated representations of data.
Critical perspectives on the role of mathematics in data-science
- Panel-session at SPT2017: The Grammar of Things. 20th conference of the Society for Philosophy and Technology
- June 5, 2017 – Darmstadt, Germany.
The upshot of this panel-discussion is to bring together different perspectives on the epistemic and societal role of mathematics in its relation to data-science and the data-revolution. It is based on the assumption that only a realistic picture of mathematics, as emphasised within the philosophy of mathematical practices, can reliably inform such an inquiry. The latter presupposes a better understanding of the role of applied mathematics in the sciences, an appreciation of the diverse ways in which statistical theory can inform the development of data-processes, and a critical outlook on the societal status of mathematics. Such a realistic picture of mathematics serves two purposes. It should inform an analysis of what it means to “trust in numbers” or help us identify clear cases of “mathwashing”, but it should just as much clarify the critical role of mathematics and explain how certain epistemic virtues of mathematics can play a decisive role in exposing epistemic failures and poor practices in data-science.
- Participants: Karen François (Brussels), Christian Hennig (UCL), Johannes Lenhard (Bielefeld), and Jean Paul Van Bendegem (Brussels).
Information Architecture and the Philosophy of Information: A dialogue
- June 27, 2017
Information is often conveyed and/or accessed across many channels, contexts, and levels of abstraction. Pervasive information architecture, a recent development in Information Architecture, is explicitly concerned with this topic and tries to develop the theoretical means to accommodate the technological changes and especially the blurring of boundaries between the digital and the physical that do not only have rendered most of our interactions with information messier than ever before (an epistemological concern), but have also radically altered the nature of the infosphere (an ontological issue).
Within the philosophy of information the possibility of accessing or shaping information in a multitude of ways is part of a broader theoretical concern, which ranges from the basic question of how and why we adopt a certain level of abstraction, to the question of how we refine, modify, compare and combine information that was obtained in different ways and at different levels of abstraction, and finally the question of how we move from obtaining information to acting on the basis of that information.
Information visualisation provides a narrow but intriguing example of how information can be accessed or at least presented across many levels of abstraction. Within the subfield of the philosophy of science concerned with modelling and representation this feature has been characterised by saying that visual representations render many possible abstractions simultaneously accessible, and put forward as one of the reasons for the effectiveness of information visualisation.
Overall, the problem of characterising pre-existing levels of abstraction is more tractable than the actual design of new levels of abstraction. The latter problem is as relevant to information architecture as it is to the philosophy of information, but the practical and theoretical resources that both fields can rely on to tackle this question are very different. On the optimistic assumption that said resources may be complementary, the question of how levels of abstraction (and their many technological implementations) are designed or engineered could be the starting point for a dialogue between the philosophy of information and information architecture.
Informational actions, or actions that alter the appearance of information, form an integral part of the design process we should focus on. As a starting point for the exchange, we propose to use the term “masking/unmasking” actions as a generic way of referring to actions that shape or alter information (including the infosphere itself and the conceptual machinery we rely on) by (and this enumeration is not meant to be exhaustive) amplifying/muting, compressing/expanding, collapsing/exploding, and merging/decomposing information, or changing the scope (include more less), scale, resolution, perspective, or context of how information is accessed, conveyed, or even made available for active intervention (modification, repurposing, co-creation). We would like to invite participants to reflect on how such actions are and should be used in practice, and in particular on how they can be used to favour, but also to negotiate and overcome trade-offs between informational virtues like, for instance, findability and legibility (core aims of IA), or precision, accuracy, truthfulness, informativeness (some very traditional aims), or “fit for purpose” (the central criterion associated with information quality, but arguably also one with a much broader appeal).