From CS 294-10 Visualization Sp10
- Sushrut Pavanaskar
- David Zats
Our project will be to demonstrate the utility of automated table design for conveying meaningful information (i.e. story) to the target audience. The software most commonly used today fails to appropriately guide the user to the appropriate representation. As a result, the user is forced to either expend significant effort or make do with suboptimal (default) table designs. This problem is exacerbated by the fact that default tables often do not follow the basic principles highlighted in existing research. For example, we have learned that avoiding gridlines and reducing non-data ink are best practices.
To resolve these issues, we plan to develop our own simple application that will take as input a csv file and ask the user a few simple questions to determine the best representation. These questions will include queries about the data, such as the type (i.e. N, O, Q). Additionally, we will attempt to determine the purpose of the table. For example, we will ask the user whether they wish to focus on trends, abnormalities, minimums, maximums, etc. We plan to devote significant effort to determining the minimum set of questions to ask the user, such that they do not become too overwhelmed with the survey. Answers to the survey questions in conjunction with the csv will be used to design and display a table that conforms with existing best practices as found in the literature today.
Initial Problem Presentation
Presentation Slides File:Fp-p1.pdf
- Link to source code and executable
- Link to final paper in pdf form
- Final Slides: File:Fp-p2.zip