Predicting Popularity

摘要

RESULTS: Researchers at Stanford University built a model that can predict, with 75 percent accuracy, when a new piece of online content's popularity will peak and how long it will last. WHY IT MATTERS: The ability to predict how widely a news story or tweet will travel could help identify the most influential blogs and Twitter posters, providing clues to who might be able to disseminate an important piece of information most broadly. Websites could use the predictions to position their content and advertising, possibly increasing click-through rates.