Prev
The most expensive political campaign season in US history is well underway, with experts predicting that spending from both the Republican and Democratic camps will exceed $5.8 billion before races for president and Congress are decided in November.
Needless to say, campaign finance is a big business. And where big money leads the way, big data is never too far behind.
That's why the Center for Investigative Reporting, the nation's largest non-profit investigative reporting organization, is teaming up with Investigative Reporters and Editors, Inc. -- the world's leading investigative and data journalism trade group -- to offer this Prospect challenge in the hopes of answering one simple question: What can some of the world's most brilliant data scientists teach us about finding hidden patterns, interesting connections and ultimately compelling stories in a treasure trove of data about federal campaign contributions?
Data journalists have been examining federal campaign finance records for decades, finding patterns and trends that have forced resignations and reforms. But despite many noteworthy successes, journalists' imaginations are limited by our skills. We're not mathematicians and machine learning experts! We might learn from a source, for instance, that a particular congressman has begun receiving campaign contributions from a new and unusual donor. But a carefully tuned anomaly detection system might reveal those patterns before our sources would ever notice them.
What kinds of ideas are we looking for with this Prospect challenge? We want to see how sophisticated clustering algorithms can help spot donors that coordinate their operations, giving to the same candidates at the same times. We want to see how classification or anomaly detection systems can find donations that are particularly interesting or unusual. We want to know how you would mix up campaign contribution data with other sources -- lobbying records, congressional votes, federal contracts -- to find patterns that journalists are missing.
Novel approaches to analysis, ideas for useful tools, and data visualizations will all be considered.
For more information, visit