eHarmony is MicroTargeting for your Lovelife
Millions of people use internet dating sites in hopes of finding that special someone, yet the actual inner workings of these sites remain somewhat mysterious. How can a mindless computer predict something as complex and personal as partner compatibility? Surely, the essence of love can never be expressed in binary.
However, innovative sites like eHarmony are getting close. Unlike some online dating tools, eHarmony doesn’t simply collect your hobbies and interests and match you with someone who input similar answers. Unfortunately, a shared appreciation for the Beatles and a golf interest does not necessitate a happy couple.
How does eHarmony do this and what does it have to do with MicroTargeting?
Back in the 90’s, eHarmony co-founder Neil Clark Warren researched over 5,000 married people, leading to the creation of a regression formula that predicts compatibility based on personality, temperament, values, and other character qualities. The formula incorporates Warren’s findings by identifying the qualities of each partner in the successful marriages he studied, and then projecting these compatibility prospects onto current eHarmony users who have corresponding personalities.
TargetPoint’s MicroTargeting shares many similarities with the eHarmony approach. We gather large amounts of individual level data and sort voters into groups according to their individual character traits (where they shop, how much they travel, their education level, etc). Then, we take their corresponding political opinions and beliefs and project them onto a larger population of people who are in the same character group.
eHarmony believes that if two people have similar personalities, values, and beliefs, they will be compatible partners. Similarly, we believe that if two people have similar personalities, values, and beliefs, they will have similar reactions to candidates, campaigns, and policy issues.
To learn more about eHarmony’s method, check out the Wikipedia article or this book about data mining featuring eHarmony.
Comments on this entry