Did you just watch the love of your life exit the L train without asking for her digits? Digital renaissance man Luke DuBois is here to help. As the Brooklyn Paper reports, Mr. DuBois, an instructor at both the Polytechnic Institute of NYU and the Brooklyn Experimental Media Center, has built a website to help the Craigslist crowd find their “Missed Connections.”
The site scrapes Missed Connections listings in nine different cities, including New York, then uses an algorithm that matches descriptive words from from listings in the same city. If it’s a close enough match (upwards of 85 percent), users are prompted to email the authors of the initial posts to let them know you may have found their subway soulmate.
So far, [Mr. DuBois] has sent e-mail alerts to eight potential couples, but he hasn’t heard back from any of them, including a particularly close match between a salami-eating man and the woman who bummed him a cigarette.
“I think you two are looking for each other,” he wrote to each of them.
Unfortunately, the wonky interface looks like something only a Linux-lover could embrace. The algorithm, which is susceptible to spam posts, is also hindered by a gender gap. “Women tend to offer more detailed descriptions of their loves at first sight, while men often keep things overly simple, with postings like: ‘You were on the R train and you were cute,’” says the Brooklyn Paper.
If nothing else, the site can be seen as an reminder that true love requires specificity. As Mr. DuBois explains:
“simple words like ‘of’ and ‘the’ have a very low score; nouns and verbs like ‘train’ and ‘kiss’ have a medium score; adjectives and adverbs like ‘blue’ and ‘softly’ have a high score; proper names like ‘kevin’ and ‘chelsea’ will score highest of all.”
“Softly”? That sounds more like “Casual Encounters” than “Missed Connections” to us.