RealViz – Interactive Visualizations for Real-Life Systems

Two novel interfaces for in-field and self-report data collection and analysis of life histories. LifeHistory interface enables direct input of multifaceted longitudinal data via a timeline grid annotated with pictorial representations of landmark events. TrajectoryView is an interactive visualization of life history data for a side-by-side comparison of parameters of an individual’s life history. The two interfaces are intended as an alternative to paper-based tools and methods currently used by qualitative researchers. Beyond a mere automation of data collection and presentation, the interfaces offer enhancements supporting recall of events and visual analysis of data. We expect that the use of LifeHistory and TrajectoryView will simplify data collection and analysis processes, leading to greater accuracy of data and better opportunities for insights.

Compared to the traditional, three-column grid-based representation of the data, the Association Map visualization is designed to make it eaier to explore data, focus on individual items and connections and, more exciting to interact with.

AM-L is an extension of the AM interface that has been enhanced with search and interaction features for supporting larger data sets. The new version provides a way to work with large data sets by employing zooming to highlight selected items, while still keeping the visualization to one screen. To facilitate easy observation of the selected items, the selected items from the middle column are also moved to the center of that column. Finally, a partial match search feature was added to facilitate searching. Displaying the same set of data in a tabular format would require multiple pages and the use of a scrolling mechanism in order to locate the needed records. [ AM 2.0 demo]

This work aims to bridge the gap between the goals of the users of information visualization systems and the techniques that are currently available to them for interacting with force-directed layouts. In our latest prototype, we have developed a new version of the prototype, in which objects can be automatically grouped based on the value of one or more properties, with each property representing a different data variable. Applying different constraint strengths to those groups provides an effective means for identifying commonalities and patterns in multivariate data.

The Dynamic Task Map (DTM) provides an interface locating a needed transaction interface via a dynamic, interactive visualization of transactions and the links between them. It is a interactive graph, derived from usage logs. The nodes in the graph correspond to transactions; links connect the active transaction with those that users typically perform in parralel or after the selected one, according to the log data. The size and coloring of each node reflects the overall frequency of the corresponding transaction. Two connected nodes are closer, when one is more likely to co-occur or follow another.

In a user study comparing DTM with a menu-based interface for locating transactions in a large enterprise system, users performed at least twice as fast with DTM and also preferred it to the menu structure. Outcomes from that study, including feedback from participants, led to improvements that were incorporated into a new version of DTM, called DTMi

Babaian, T., Lucas, W., Chircu A. and Power N. Evaluating Interactive Visualizations for Supporting Navigation and Exploration in Enterprise Systems. ICEIS (2) 2016: 368-377 [pdf] [demo]

The Dynamic Task Map with information, or DTMi, is an extension of the DTM interface. DTMi extends the original DTM interface by introducing (a) grouping of tasks by functional module within distinctly colored convex hulls, (b) a dedicated panel on the right side for displaying additional task information that is not communicated by the graph itself, (c) an improved search interface.

A design pattern library serves two major purposes. First, it represents design knowledge about interactive visualization for real-world systems, and second, it provides a context-dependent way for the designer to explore and search the library to find the appropriate design patterns. We aim to improve on the existing models of organizing a design pattern library by providing an adaptive mechanism that takes into consideration various contextual elements, such as problem to be solved, context, examples of solutions, reasons why solution is appropriate, and related design patterns. In our model, each design pattern is internally described using ontological relations similar to the semantic web. Using these relations and the designer’s behavioral pattern, the library can compute in a flexible and context-specific.