Particle Physics Improves Search Engine Recommendations
Particle physics may be about to improve your social life. Researchers have discovered similarities between how particles behave and recommendations of venues and services made by Internet search engines.
We’re all familiar with pop ups and ads telling us that our recent purchase of one product suggests that we’ll like another, similar product. These recommendations are an attempt to pair each consumer with a specific product or service. However, what’s never been studied is what effect making the same recommendation to many people will have on the user experience, said Matus Medo, a research assistant at the University of Fribourg in Switzerland.
The dilemma is clear: If a restaurant, bar, or other activity is too highly recommended, the user experience may be diminished. For example, if a quiet and romantic restaurant is recommended to many people, and all those people choose to go to the restaurant, it will no longer be quiet and romantic. The wait times for tables and for food will also increase. But for other types of recommendations, such as books or music, the number of recommendations doesn’t matter. Multiple people can read the same book or listen to the same song without getting in each other’s way.
“In some situations it’s very favorable to avoid other users and in other situations you simply don’t care,” Medo said. “In particle physics you also have particles which try to avoid others but you have other particles which do not care about sharing the same space.”
In physics, particles tend toward energetically favorable states. Some particles, such as bosons, don’t have limits to how many can occupy the same state. Others, such as fermions, are unable to occupy the same state in large numbers. Sometimes, having just two particles in the same state is too much.
In addition to finding a parallel between particle physics and recommendation search engines, Medo and his team also found that the incorporation of crowd avoidance to search engine recommendations leads to better recommendations. In other words, recommending a restaurant because it is less crowded, and not because it is the best fit for a consumer, can actually create a superior user experience.
All that being said, until search engines start incorporating crowd avoidance, opt for your second choice restaurant, bar or concert. “We’re taking a wholly new perspective on recommendations,” said Medo.
New recommendations may not be the best fit, but in the end you may just have a better time.