It had to happen. Computer scientists have just come up with a way to predict before publication whether or not a news story has the potential to create buzz.
MIT bloggers wrote on a “Physics arXiv Blog” post this week that computer scientists at HP’s lab in Palo Alto, Calif. did research that showed they could test news articles to see whether or not they would spread widely on Twitter. Since Twitter buzz about news stories has been shown to be a predictor of general interest beyond Twitter, the ability to
test an article in advance could have major implications for journalism, and, of course, for public relations as well.
The HP scientists used an automatic online news aggregator to collect news stories for a week. Then they scored each story on four different criteria: what news organization published the story it, what category the story fit into, how subjective the language used in the story was, and what people or things were included in the story. They tracked these stories on Twitter to see how far and how fast they were spread. They were able to use the data to identify what levels of scores in each of the four criteria were correlated with the popularity of a news story on Twitter. Then they used these criteria and their scoring formulas to predict in advance how popular a story would be on Twitter.
The MIT bloggers speculated that it probably wouldn’t be long before someone would use this type of data to develop a “popularity checker” tool, similar to the grammar and spelling checkers that are built into word processing programs. They commented that it might be detrimental to journalism to have such a tool, since journalists would surely be pushed by their employers to write for the tool.
But think of what a boon to PR a popularity checker would be!
PR industry newbies would be able to test their news sense to get objective feedback on their news releases and media pitches. PR team leaders could use the checker to demonstrate tactfully to their employers or clients that promotional language really does not work in a news release, or to point out the lack of newsworthiness of an announcement proposed by senior management without having to argue about it.
If anyone hears that such a tool is under development, please let me know so I can sign up as a pre-release beta tester and invest in whatever company is behind it!