With the ever-growing amount of data produced by humans as well as IoT devices and smart environments, there is a strong need for tools and natural user interfaces to explore and analyse these large datasets forming part of the Big Data era. There has been a lot of research on how to efficiently batch process large datasets and create some static reports for the analysis of the underlying dataset, but recently there is an increasing interest in real-time processing and query adaptation in Big Data environments, enabling so-called human-in-the-loop data exploration. While the majority of user interfaces for data exploration and analysis are based on traditional screen-based visualisation, the emerging field of data physicalisation offers new opportunities since data can no longer only be explored visually, but also by making use of other senses such as touch. This new form of cross-media interfaces further enables the multi-user and real-time interaction and exploration of datasets since there is no longer a single user in control of the interface via keyboard and mouse. Dynamic data physicalisation enables innovative forms of higher-dimensional data representation where various physical properties (e.g. temperature or tactile feedback) can be combined with a 3D data visualisation.

The proposed research agenda will address a number of challenges on the data processing side, such as special types of dynamic index structures, as well as on the data physicalisation side (e.g. design guidelines for the use of physical variables or multi-user interaction) to realise real-time cross-media data exploration and analysis for next-generation knowledge workers. The envisioned data physicalisation framework will enable other researchers to conduct their dynamic data physicalisation research or studies but also serve as a platform for the rapid prototyping of dynamic data physicalisation solutions for ordinary users. Further, an algebra defining the supported operators will serve as an interface between data processing and data physicalisation. There is a broad range of activities that might be supported by the proposed innovative real-time cross-media data exploration interfaces, ranging from the analysis of domain-specific research data to the exploration of large datasets by regular users (e.g. datasets collected in smart city management where real-time pollution or traffic data might be analysed). Last but not least, the proposed real-time explorative human-information interaction might also be used during leisure activities such as museum visits.

Related Publications