Typical matchmaking algorithms for carpooling use geographic and temporal data. University of Waterloo researchers have found that incorporating the social aspects of the carpoolers in matchmaking algorithms can make ridesharing a success, and significantly reduce vehicle usage. The study used social media analytics, algorithms and computer simulations to arrive at the conclusion, and the results have been published in Transportation Research Part C.

Bissan Ghaddar, author of the study says, “Usually carpooling is about just matching people depending on geographical location and time of schedule. We wanted to include the social aspect into the equation, because it’s always awkward when there is silence in the car, especially if it’s a long commute. We believed that we really needed to look at the social aspect, and our initial data analysis agreed with us.”

The Twitter feeds of potential carpoolers was analysed to identify the interests of the people involved. The followers of the Twitter users were also examined to check if the potential carpoolers engaged more with familiar friends or with strangers. The results were fed to a matchmaking algorithm, and then the results of the matchmaking were run through a computer simulation using real world data. The results showed that the carpoolers were more satisfied with the ride, and car use reduced by 57 percent in Rome and 40 percent in San Francisco.

Publish date: July 5, 2017 7:33 pm| Modified date: July 5, 2017 7:33 pm

Tags: , , , , ,