Data Analysis of OkCupid Profiles

Posted on May 1, 2017
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.

Online dating sites are an increasingly popular way to find a mate. With 17% of marriages sprouting from an online dating service, data shows the formerly desperate activity has lost much of its stigma. More and more people are uploading profiles to attract a special someone. What insights can these curated self-portraits provide?

The Data

The OkCupid profiles dataset contains information from 59,946 San Francisco users. This dataset was scraped from active profiles containing at least one photo in June 2012 and made available to the public. Each profile contains information on gender, sexual orientation, ethnicity, religion, education, physical features, and more for a total of 31 possible features. Income, height, and age are the only available numeric features, making much of the data categorical. Ten possible essays were available as well, answering prompts such as "Describe your typical Friday night" and "What is the first thing people notice about you?"

OkCupid allows the user to have fun with their profile and present a unique version of themselves, making the data messy. For some features, such as ethnicity and offspring, users are allowed to pick multiple responses, making the data messy.

Cleaning the Data

My first step was to clean this up so I could analyze the population by groups. I focused on ethnicity, religion, astrological sign, offspring, and pets. Observations with multiple ethnicities listed were categorized as "multiethnic" unless the only additional ethnicity was "other", in which case the ethnicity was stemmed to the primary category. Religion and astrological signs were separated into two columns each: one affiliating the user to their selected religion and sign, and the other noting the user's indicated level of devoutness to religion or astrology. Pets was split into a cats and dogs category, while offspring was split into "has kids" and "wants kids".

For certain categories, many people opted to keep information private. Β In particular, OkCupid users seemed reluctant to list their income and whether or not they have or even want children.

Data Analysis of OkCupid Profiles

Data Exploration and Visualization

From the available data, many interesting relationships can be seen. First, we can look at the typical male and female by aggregating the most common responses for each category. Outnumbering women 3:2 in this dataset, the typical male is one year older than Β his female counterpart. Much remains the same across gender lines, except that men seem to be taller.

Data Analysis of OkCupid Profiles


When looking at averages, however, the salary gap can be seen with mean making $24,351 more than women. The graph below shows the distribution of salary among men and women. Β This disparity existed despite the fact that women on the site were more likely to have a masters degree or be in law or med school.

Data Analysis of OkCupid Profiles



Word Clouds

Differences can also be seen in the way men women wrote about themselves. In a word cloud analysis of the most commonly used items that men and women "just cannot live without," it seems that women care more about love, books and coffee while men enjoy the internet. Both groups like good food and good music.


Some men and women listed their highest level of education as space camp. Interestingly, this group showed significant differences from the general population. Those who listed space camp were 23% likely to sometimes use drugs compared to the average of 13%. Females in particular were twice as likely to not want kids if they claimed to have gone to space camp.

Another interesting group was bisexuals. Although only making up 5% of the users, this unique group showed significant differences in ethnic and gender makeup, with far more women describing themselves as bisexual than men. As shown below, they were also more likely than gay and straight people to be agnostic or atheist or ascribe to some "other" belief system. According to the data, this group was also more likely use drugs and own cats.



Conclusions and Further Questions

Although this data proved fun and interesting, it offers limited business insights. The goal of OkCupid is to subscribe and match users to spark a potential romance. The profiles do not have any information on communication between users whatsoever, so we can't conclude if a particular response or non-response correlates with higher levels of successful interaction with other users.

We can say, however, that based on the ethnic and gender makeup of the profiles, OkCupid is missing out on potential members in the San Francisco area. Asians, Hispanics, women, and gay people are particularly underrepresented on the website. However, younger age groups are more diverse than older age groups, indicating this inadequate representation may change. Further research should still be conducted. Other interesting topics to explore would be message and reply rates by certain groups and why some users fail to stick with the site or even complete a profile.

About Author

Sean Kickham

Sean migrated from the Midwest to New York City after graduating with a BS in Mathematics from the University of Notre Dame. He taught middle school math for five years in city schools. Equipped with a Masters in...
View all posts by Sean Kickham >

Related Articles

Leave a Comment

Sean Kickham October 24, 2017
This was a previously-collected dataset I found online, and so must be a couple of years old when the income data was available.
John Danielson October 24, 2017
Where were you pulling income data from? The site hasn't had that on profiles in over two years...

View Posts by Categories

Our Recent Popular Posts

View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI