This paper was interesting to read because it presents a list of tasks for graph visualization which could be used by designers as well as evaluator. Just by reading the abstract I could figure out that it might talk about something which we could use in our class project. Although there was not much at the implementation level but it did put forward some theoretical terms and classifications.
A good thing about the paper was that they provided benchmark data set to support their point and hence improve the evaluation of info vis system. Another point of interest was that they did not assume anything about the knowledge set of their audience and thus present everything from the scratch. They defined all the graph-related objects like nodes, links, paths, clusters, sub-graphs etc.
After defining the definition for graph components they presented a list of low-level visual analytic tasks relevant to graphs. Some of them included retrieve values, filter, sort, find range, cluster, correlate. If you really think about these tasks they are really very basic and I don't think they can be further broken down into sub-tasks. So I had to agree with the authors on this.
Moving forward they categorized the above specified tasks into four different groups based on topology, attribute, browsing and overview. The best part of the paper was that they provided general description and examples scenarios for each of these categories, which made it really ease to understand what they were trying to say.
In the end they also talked about high level tasks but I don't think they were really trying to focus on that as the discussion was very brief.
0 comments:
Post a Comment