Team gravite

Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: New Results

User Evaluation

Participants : Daniel Archambault, Bruno Pinaud.

The ACM special interest group on Computer-Human Interaction (HCI) “ACM SIGCHI Curricula for Human-Computer Interaction” proposes a definition of HCI (see ):

Human-computer interaction is a discipline concerned with the design, evaluation and implementation of interactive computing systems for human use and with the study of major phenomena surrounding them.

We already work in the design and implementation of interactive techniques for visualization and interaction with graphs. More generally, many techniques and systems have been proposed in the InfoVis community but there is a lack of user evaluation of those systems. Thanks to close collaboration with Daniel Archambault (lique Strategic Research Cluster at UCD Dublin) and Helen Purchase (Department of Computing Science, University of Glasgow ) we did three differents user evaluation. The data sets used in the experiments were derived from standard benchmark data sets of the information visualization community. The Questions used were selected to test both local and global properties of the graphs.

Animation, Small Multiples and the Effect of Mental Map Preservation in Dynamic Graphs

The majority of dynamic graph drawing algorithms have been designed with an animation of the sequence of graphs as their output in mind. In the information visualization community, small multiples has been in use for many years to display dynamically evolving data. In a small multiples approach, a matrix of images shows the differences between objects. In the case of dynamic data, each image in the matrix is a timeslice. Users are able to see and compare all timeslices at the same time.

We compared the performance of the animation of dynamic graphs to the presentation of small multiples and the effect that mental map preservation had on the two conditions. We found that small multiples gave significantly faster performance than animation overall and for each of our five graph comprehension tasks. In addition, small multiples had significantly more errors than animation for the tasks of determining sets of nodes or edges added to the graph during the same timeslice, although a positive time-error correlation coefficient suggests that, in this case, faster responses did not lead to more errors. This result suggests that, for these two tasks, animation is preferable if accuracy is more important than speed. Preserving the mental map under either the animation or the small multiples condition had little influence in terms of error rate and response time.

More details can be found in [12] .

The readability of Path-Preserving Clusterings of Graphs

Graph visualization systems often exploit opaque metanodes to reduce visual clutter and improve the readability of large graphs. This filtering can be done in a path-preserving way based on attribute values associated with the nodes of the graph. Despite the extensive use of these representations, as far as we know, no formal experimentation exists to evaluate if they improve the readability of graphs. We ran a user study that formally evaluates how such representations affect the readability of graphs. We also explore the effect of graph size and connectivity in terms of this primary research question. Overall, for our tasks, we did not find a significant difference when this clustering is used. However, if the graph is highly connected, these clusterings can improve performance. Also, if the graph is large enough and can be simplified into a few metanodes, benefits in performance on global tasks are realized. Under these same conditions, however, performance of local attribute tasks may be reduced.

More details can be found in [18] .

Difference Map Readability for Dynamic Graphs

Difference maps are one way to show changes between timeslices in a dynamic graph. They highlight, using colour, the nodes and edges that were added, removed, or persisted between every pair of adjacent timeslices. Although some work has used difference maps for visualization, no user study has been performed to gauge their performance. We did a user study to evaluate the effectiveness of difference maps in comparison with presenting the evolution of the dynamic graph over time on three interfaces. We found evidence that difference maps produced significantly fewer errors when determining the number of edges inserted or removed from a graph as it evolves over time. Also, difference maps were significantly preferred on all tasks by our users.

More details can be found in [17] .


Logo Inria