Bernardo A. Huberman: Social attention in the age of the web

Social attention in the age of the web

- How to replace the editor with a computer

By Bernardo A. Huberman, Senior Fellow and Director

Information Dynamics Lab, HP Laboratories, Palo Alto, California

IT is the ultimate editorial decision: what to put on the front page and where to put it. Should pride of place go to another piece on the presidential election, because that is what everyone is excited about? Or would a story about a parrot that can do algebra be more eye-catching, because that is what nobody is expecting?

Editors usually make their decisions based on simple rules of thumb, such as how many days in a row the elections have been on the front page and what subjects other newspapers are focusing on, as well as their gut feelings about whether readers will be intrigued by, say, a novel animal story.

On the web, though, the competition between popularity and novelty takes on a new dimension, because it is easy to change the choice and line-up of stories many times a day, even many times an hour. It is also easy to measure which stories are getting the most attention.

So Fang Wu and Bernardo Huberman, a pair of researchers at Hewlett-Packard's laboratory in Palo Alto, California, decided to compare two strategies designed to maximise readership, one based on the previous popularity of a story, the other on its novelty. What they found is that the best strategy depends, quite sensitively, on how quickly readers tire of a new story—a result that could turn editorial decisions into a rational process, rather than an intuitive one.

Organized by Center for Applied Information and Communication Technology (CAICT)