Talk:Visualizations
From Business Intelligence
Who is using such visualizations for University work now? If you aren't including even simple graphics into your reports, why aren't you? Is it that no one is asking for charts? Is it that it takes extra time? Is it that you don't know the right type of visualization to represent your data effectively? Is it that you need to have repeatability and it's your experience that adding graphs, etc is extra time that requires custom work?
Thanks in advance for the time and thought that will go into your responses.
Mark
[edit] Advice sought for data that's too complex for a Venn Diagram (24 circles)
I have 20,000 rows of two columns of data I'd like to visualize. It's a list of uniqnames and each data warehouse data set the uniqname has access to.
johndoe CRASSEL johndoe DACSEL janedoe DACSEL janedoe M_FADW1_DEF_SEL
I'm thinking Venn diagram, but it'll wind up too messy because there are about 24 data sets I want to show. I could do one of those word bubble things where the words Fin and Payroll are much bigger than the words "Student Records" and "DAC" because six times as many people have access to the former. But the intersections are important to me, as are the counts of people with access to a single data set.
Any advice appreciated. -Mark
Responses
I suggest a heat map in the form of a matrix, where the rows and columns are the data sets and the intersections have a circle whose size or color relate to the nuber of users who have access to both data sets with a total row at the botom.
I would also do a summary table with the first column listing the data sources, the next column showing the number of users with access, and the last column with the number of other data sets it shares users access with. Some in-cell micrographs might aid with scaning such a table.
TJP 09:48, 6 November 2008 (EST)