The first few sections in the paper attempt to identify when InfoVis can be utilised to present a data set. As stated by the paper, I agree that before visualizations are to be conceived, emphasis have to be given to the target audience and their general “visual literacy”. I feel the relativity in cognition derived from the InfoVis amongst the population is a challenge along with devising methodologies to teach as to when InfoVis can be useful. The paper goes on to state arguments favouring perception‘s processing capabilities and cites examples that unearthed knowledge otherwise hard to perceive. I agree that visualization speeds up “filtering amongst competing theories”. However, optical illusion, characterized by visually perceived images that differ from objective reality, can creep in leading to erroneous conclusions. Hence, visualizations have to be planned and experimented upon extensively. Indeed, experimentation is one such cost mentioned by the parameter Ci (research costs) in the paper.
The paper also makes comparisons between InfoVis and automatic analysis. The authors classify data mining and InfoVis separately as automated and exploratory methods. I would like to comment that an InfoVis can be a result of extensive automatic analysis that can help complement the knowledge derived after the mining process. Hence InfoVis and auto analysis can have a derivative relationship. Finally an attempt is made in the paper to derive an economic model to quantify value. A parameter that can be added here,to reflect long term value is maintenance costs. We live in a dynamic world, where data and thus their visualizations are ever changing.
0 comments:
Post a Comment