I worked at MicroStrategy, a software company that develops data-mining software to aid in database analysis and decision support, from March 2000 through May 2001. MicroStrategy's products capabilities provide the ability to analyze vast amounts of data and gain valuable business insight.
As an example, perhaps you are in charge of a political campain. Let us further assume (this would be a stretch) that you had a database containing the the way every person voted in the last, oh how about 4 national elections. You also know a lot of demographics about those people. (Big brother is watching) With this impossible theoretical situation let us see some of the questions that our software could help you answer.
- Rank all major US cities according to percentage change in voter favorability after a city visit.
- Rank by income, location, age and nationality those voters who changed their partyline voting characteristics between the last 2 elections.
- Find the correlation between the votes of different states. For those with high correlations, which primaries were first?
- To what extent did women who'd had an abortion between elections alter their voting paterns relative to the candidates message?
- What percentage of young adults voting for the first time vote along the same part lines as their parents?
- Assuming that percentage is constant, what percentage impact will those incoing voters have in the next election.
This is but a sampling. If the data exists and the relationships can be define in a database, we can pull calculated knowledge from the raw data.
