I've implemented a data mart based on campaign finance data from the Federal Election Commission (FEC) and will be blogging about the project from two perspectives:
- Leading up to the 2008 elections, I'll post the results of my analysis on an ongoing basis. I hope this will be of interest to the general public.
- Of potential interest to fellow SQL Server practitioners, I'll post various MDX and SQL code samples, and discuss the technical challenges I encounter.
The FEC web site allows the public to query this data based on a number of parameters. However, the query results are incredibly flaky, which provided the original inspiration for this project. (And now that I'm acquainted with the data, the flakiness comes as no surprise -- the data is dirty. I've done some cleaning but it's work in process.)
I discovered there are a handful of other sites that have embarked down this same path. However, I have found the reports they publish to be lacking, hence this project.
The relational data mart is implemented using SQL Server 2005, with an Analysis Services cube sitting on top. If a sponsor should emerge (hint, Microsoft), I'll make the database available for online querying.