ABSTRACT
Computational news discovery refers to the use of algorithms to orient editorial attention to potentially newsworthy events or information prior to publication. In this paper, we describe the design, development, and initial evaluation of a computational news discovery tool called Lead Locator, which is geared towards supplementing national politics reporting by suggesting potentially interesting locations to report on. Based on massive amounts of data from a national voter file, Lead Locator ranks counties based on statistical properties such as their extremity in the distribution of a variable of interest (e.g. voter turnout) as well as their political relevance in terms of shifts in voting patterns. It then presents an automatically generated tip sheet of potentially interesting locations that reporters can interactively browse and search to help inform their reporting ideas.
Algorithms surface leads for U.S. political coverage in another development from the emerging field of computational news discovery. One algorithm assesses massive sets of U.S. political data to determine anomalies in voter behaviours, then another turns the results into readable text as a tip sheet.
The tool turns up leads otherwise unnoticed by human reporting.
They call their tool Lead Locator. It was designed by the engineering group at The Washington Post in collaboration with reporters and editors in the Post newsroom.
Other algorithms for computational news discovery are used to monitor emerging anomalies or clusters, for example, by Reuters to look for news alerts by watching Twitter (called Tracer).
The lead author of this paper describing the project is Nick Diakopoulos, a professor at Northwestern University and author of the leading book about journalism and algorithms, Automating the News: How Algorithms Are Rewriting the Media.
‘Lead Locator uses data mining to analyzes a national voter file tracking every registered voter in the U.S. in order to rank counties based on their potential newsworthiness to reporters.’
AUTHORS
- Nicholas Diakopoulos – Northwestern University
- Madison Dong – Northwestern University
- Leonard Bronner – The Washington Post
- Jeremy Bowers – The Washington Post