We explain the objectives of our research program by referring to our recent publication in this area.

Jacob Cybulski, Susan Keller, Lemai Nguyen and Dilal Saundage: Creative problem-solving in digital space using visual analytics, Computers and Human Behavior, V42 (January), 2015, pp 20-35. Excerpts.

Interactive visual analytics (IVA) is a visual approach to analytical reasoning, which has set in motion a paradigm shift for handling massive, open, and complex data. IVA combines human perception, cognition, and interaction with analytics to process massive data sets. Our society is truly immersed in the digital world. Digital “natives” generate and consume a vast amount of data in a variety of formats at breakneck speed. For example, the amount of data generated annually grew from 150 exabytes in 2005 to 1200 exabytes in 2010 and it is predicted to grow by 40% every year in the near future. The huge quantity of readily available data, which is also referred to as “big data”, comes in many different forms including text, web data, tweets, sensor data, audio, video, click streams and log files from sensors, smart devices, and social collaboration technologies. Electronic instrumentation and web connectivity continually generate data at a speed that conventional systems can no longer capture, store, and analyse.

The worldwide data explosion raises a key challenge as to how to make sense of big data. This problem requires novel analytical approaches to examining big data, especially because much of the newly generated data is not well understood. Large parts of freely available data are unstructured and often describe human activities that are inherently unpredictable. These data are difficult to analyze with simple graphing and non-interactive techniques. It is also a challenge to analyze relationships between data elements that are not explicit and often not known up front. The very nature of big data provides a strong case for the pursuit of novel analytical approaches capable of creatively combining formal data analysis, rich data visualization, use of familiar metaphors for data representation, and interactive manipulation of these representations, leading to comprehensible reports.

IVA allows analysts to combine computational power of computers with additional perceptive and cognitive abilities to interactively manipulate and derive insights in support of decision-making from big data. The analysts’ engagement with data visualisations plays a significant role in the discovery and communication of insight in IVA. Such engagement involves iterative analysis of datasets from a variety of different viewpoints and interactive manipulation of data attributes to gain new knowledge that is not readily apparent. In other words, IVA aims to detect the expected and discover the unexpected from massive, dynamic, ambiguous, and often conflicting data.

IVA has its roots in the fields of scientific and information visualisations. Scientific visualisation focuses on the study of large pools of scientific data from sensors, simulations or laboratory tests; while information visualisation deals with communication of abstract data and providing the capabilities to transform this data through the use of interactive interfaces. In the past, neither of these visualisations dealt specifically with decision-making, whereas IVA uses decision sciences and analytical sciences to directly assist in decision-making processes based on insights derived from the analysis of big data.

The main premise of IVA is that interactive visualisation provides a more effective way to understand and discover new insights from big data, especially in the process of making significant decisions or formulating data-driven action plans.

Leave a Reply