Name matching in the Lemonade Aftermath
How wrong was the “Beyhive” when they mistook Rachael Ray for Rachel Roy?
True to form, Beyoncé once again “broke the internet” Saturday night with the surprise drop of her sixth solo album and accompanying visual album of the same name, Lemonade. Typically a very private celebrity, Beyoncé shocked fans with the album’s themes of infidelity and marital strife. The seemingly personal content of the album has had fans–self named, the “Beyhive”–buzzing ever since with speculation on the health of the starlet’s marriage to rapper Jay-Z, and the identity of the other woman: “Becky with the good hair.”
Initially premiering on HBO Saturday night, Lemonade is exclusively available via TIDAL, Jay-Z’s music streaming service. Despite TIDAL editor Ryan Pinkard calling the album “a conceptual project based on every woman’s journey of self knowledge and healing,” the internet hordes are convinced Beyoncé’s lyrics are a window into a dark period in the power couple’s marriage, and immediately set out to identity the woman responsible. Social media chaos ensued.
When fashion designer Rachel Roy posted the picture below with the caption “Good hair don’t care” shortly after the album dropped, the Beyhive considered it an admission to being “Becky with the good hair,” and crucified her on social media with posts, comments on her pictures, and even attacks on her family’s social media accounts. (Roy has since issued a public statement denying any involvement and taken down the picture.)
Unfortunately, some angry fans mistook Rachel Roy, the fashion designer, for Rachael Ray, the Food Network celebrity chef, and swarmed the latter’s social media accounts with angry posts and comments as well. Once the misunderstanding was sorted out Ray lightheartedly posted a recipe for lemonade, and these fans became the punch line of a number of internet jokes.
But just how wrong were Beyoncé fans when they confused Rachel Roy and Rachael Ray? From a purely linguistic perspective which doesn’t consider the pop culture context, only 16.4%.
Among its suite of text analysis tools, Rosette® Text Analytics powers a name matching function which compares two entity names (person, location, or organization) across 15 different languages and scripts, and provides a match score of how likely it is that these two names refer to the same person. Using the Rosette API, Roy and Ray are an 83.6% match, a respectably high match score. Compare that to Papageno and Papagena, two distinct characters from Mozart’s The Magic Flute, who have a match score of 89.9%.
To conjecture a bit then, this means that if Rachel Roy were on a TSA watch list, Rachael Ray would likely get flagged for additional screening while passing through airport security. Granted, the TSA would verify that Ray is the same person as Roy before
accusing her of having an affair with Jay-Z keeping her off her flight, but you can see how the confusion could occur.
If you thought the similarity in those two names was difficult, consider how someone might have to deal with these scenarios:
Real world implications
Outside of this Bey-bubble, matching names is a very real problem that has far more serious repercussions than a few nasty tweets. If you’ve heard of the term “Sharing economy” it refers to the likes of Uber and Airbnb, companies that are revolutionizing their respective industries, where user reviews and background checks are not enough to manage the risk involved in trusting strangers for a ride or a place to stay.
In response, Airbnb built their Verified ID feature, which verifies online IDs such as Facebook, LinkedIn, and Google+ against passports and other government-issued IDs, Airbnb provides both hosts and travelers with unprecedented trust in the transaction.
Want to learn more about name matching, or compare names yourself? We’ve gone ahead and pre-filled the live demo with Rachel/Rachael, but you can enter any two names you’d like for an immediate score. Bonus: we’ve also updated our entity extraction demo with an article on the Beyhive’s latest (how did Rita Ora get into the mix?!). Check it out to see the key players in this internet drama, or enter the URL of another article you’d like to analyze into the demo.
Rosette Text Analytics is available as an API or SDK, bringing meaning to your unstructured text data for over 20 years. Try it out today! In the meantime, cut the Beyhive a little slack. They’re just overzealous in the defense of their queen bee.