Deep User Models

Deep User Models for Visual and Immersive Analytics

---

Deriving insight from data is instrumental to any industry and government organisation for informed decision making and accelerated business agendas. The path from Data-to-Insight (D2I), however, is typically cumbersome involving tedious manual handling of data, complex analytics and often misleading visualisation techniques. Visual and immersive analytics alleviates this issue by integrating computational and human efforts, allowing for effective data exploration through interactive user interfaces. The challenge remains to optimise and automate the visual analytics process through specification of the D2I ecosystem, incl. data characteristics, system properties, analytics processes, and visualisation techniques, among other factors.

In this PhD project, we design and integrate Deep User Models (DUMs) into the visual and immersive analytics process with the goal to optimise user performance and experience for a breadth of end users. The DUMs include explicit knowledge about the user (e.g. expertise, skills, domain knowledge) as well as implicit knowledge (e.g. preference for certain information or visualisation types). The former can be specified in user profiles whereas the latter can be learned from the interaction of the user with the system.

Investigators

Amit Jena (Lead), PhD Candidate at School of Design, Indian Institute of Technology Bombay
Prof. Venkatesh Rajamanickam, School of Design, Indian Institute of Technology Bombay
A/Prof. Tim Dwyer, Computer Human Interaction and Creativity, Monash University
Dr. Ulrich Engelke, Decision Sciences, CSIRO Data61
Dr. Cecile Paris, Decision Sciences, CSIRO Data61

Lifetime

2017 – 2020