Unit9 EHarmony AllSlides
-
Upload
arun-thomas -
Category
Documents
-
view
231 -
download
0
Transcript of Unit9 EHarmony AllSlides
-
7/24/2019 Unit9 EHarmony AllSlides
1/25
eHarmonyMaximizing the Probability of Love
15.071x The Analytics Edge
-
7/24/2019 Unit9 EHarmony AllSlides
2/25
About eHarmony
15.071x eHarmony: Maximizing the Probability of Love 1
Goal: take a scientific approach to love and marriageand offer it to the masses through an online datingwebsite focused on long term relationships
Successful at matchmaking
Nearly 4% of US marriages in 2012 are a result ofeHarmony
Successful business Has generated over $1 billion in cumulative revenue
-
7/24/2019 Unit9 EHarmony AllSlides
3/25
The eHarmony Difference
15.071x eHarmony: Maximizing the Probability of Love 2
Unlike other online datingwebsites, eHarmony does nothave users browse others
profiles
Instead, eHarmony computesa compatibility score betweentwo people and usesoptimization algorithms todetermine their users bestmatches
-
7/24/2019 Unit9 EHarmony AllSlides
4/25
eHarmonys Compatibility Score
15.071x eHarmony: Maximizing the Probability of Love 3
Based on 29 different dimensions of personality includingcharacter, emotions, values, traits, etc.
Assessed through a 436 question questionnaire
Matches must meet >25/29 compatibility areas
-
7/24/2019 Unit9 EHarmony AllSlides
5/25
Dr. Neil Clark Warren
15.071x eHarmony: Maximizing the Probability of Love 4
Clinical psychologist who counseled couples andbegan to see that many marriages ended in divorce
because couples were not initially compatible
Has written many relationship books: Finding the
Love of Your Life, The Triumphant Marriage,
Learning to Live with the Love of Your Life and
Loving It, Finding Commitment, and others
-
7/24/2019 Unit9 EHarmony AllSlides
6/25
Research!Business
15.071x eHarmony: Maximizing the Probability of Love 5
In 1997, Warren began an extensive research projectinterviewing 5000+ couples across the US, which
became the basis of eHarmonys compatibility profile
www.eHarmony.comwent live in 2000
Interested users may fill out the compatibility quiz, but
in order to see matches, members must pay amembership fee to eHarmony
-
7/24/2019 Unit9 EHarmony AllSlides
7/25
eHarmony Stands Out From the Crowd
15.071x eHarmony: Maximizing the Probability of Love 6
eHarmony was not the first online dating website andfaced serious competition
Key difference from other dating websites: takes aquantitative optimization approach to matchmaking,
rather than letting users browse
-
7/24/2019 Unit9 EHarmony AllSlides
8/25
Integer Optimization Example
15.071x eHarmony: Maximizing the Probability of Love 1
Suppose we have three men and three women
Compatibility scores between 1 and 5 for all pairs
1
354
22
1
5
3
-
7/24/2019 Unit9 EHarmony AllSlides
9/25
How should we match pairs together to maximize
compatibility?
1
354
22
1
5
3
Integer Optimization Example
15.071x eHarmony: Maximizing the Probability of Love 2
-
7/24/2019 Unit9 EHarmony AllSlides
10/25
Data and Decision Variables
15.071x eHarmony: Maximizing the Probability of Love 3
Decision variables: Let xijbe a binary variable taking value 1 ifwe match user iand userjtogether and value 0 otherwise
Data: Let wijbe the compatibility score between user iandj
1
354
22
1
5
3
-
7/24/2019 Unit9 EHarmony AllSlides
11/25
Objective Function
15.071x eHarmony: Maximizing the Probability of Love 4
Maximize compatibility between matches:max w11x11+ w12x12+ w13x13+ w21x21++ w33x33
1
354
22
1
5
3
-
7/24/2019 Unit9 EHarmony AllSlides
12/25
Constraints
15.071x eHarmony: Maximizing the Probability of Love 5
Match each man to exactly one woman:x11+ x12+x13= 1
1
354
22
1
5
3
-
7/24/2019 Unit9 EHarmony AllSlides
13/25
Constraints
15.071x eHarmony: Maximizing the Probability of Love 6
Similarly, match each woman to exactly one man:x11+ x21+x31= 1
1
354
22
1
5
3
-
7/24/2019 Unit9 EHarmony AllSlides
14/25
Full Optimization Problem
15.071x eHarmony: Maximizing the Probability of Love 7
max w11x11+ w12x12+ w13x13+ w21x21++ w33x33
subject to: x11+ x12+x13= 1
x21
+ x22
+x23
= 1
x31+ x32+x33= 1
x11+ x21+x31= 1
x12+ x22+x32= 1
x13+ x23+x33= 1
x11, x21, x31, x12, x22, x32, x13, x23, x33are binary
Match every man with
exactly one woman
Match every woman
with exactly one man
-
7/24/2019 Unit9 EHarmony AllSlides
15/25
Extend to Multiple Matches
15.071x eHarmony: Maximizing the Probability of Love 8
Show woman 1 her top two male matches:x11+ x21+x31= 2
1
354
22
1
5
3
-
7/24/2019 Unit9 EHarmony AllSlides
16/25
Compatibility Scores
15.071x eHarmony: Maximizing the Probability of Love 1
In the optimization problem, we assumed thecompatibility scores were data that we could inputdirectly into the optimization model
But where do these scores come from?
Opposites attract, then they attack Neil Clark Warren
eHarmonys compatibility match score is based onsimilarity between users answers to the questionnaire
-
7/24/2019 Unit9 EHarmony AllSlides
17/25
Predictive Model
15.071x eHarmony: Maximizing the Probability of Love 2
Public data set from eHarmony containing featuresfor ~275,000 users and binary compatibility results
from an interaction suggested by eHarmony
Feature names and exact values are masked to protect
users privacy
Try logistic regression on pairs of users differencesto predict compatibility
-
7/24/2019 Unit9 EHarmony AllSlides
18/25
Reduce the Size of the Problem
15.071x eHarmony: Maximizing the Probability of Love 3
Filtered the data to include only users in the Bostonarea who had compatibility scores listed in the dataset
Computed absolute difference in features for these1475 pairs
Trained a logistic regression model on these
differences
-
7/24/2019 Unit9 EHarmony AllSlides
19/25
Predicting Compatibility is Hard!
If we use a low thresholdwe will predict morefalse positives but also
get more true positives
Classification matrix forthreshold = 0.2:
Model AUC = 0.685
15.071x eHarmony: Maximizing the Probability of Love 4
Act\Pred 0 1
0 1030 227
1 126 92
-
7/24/2019 Unit9 EHarmony AllSlides
20/25
Other Potential Techniques
15.071x eHarmony: Maximizing the Probability of Love 5
Trees Especially useful for predicting compatibility if there are
nonlinear relationships between variables
Clustering
User segmentation
Text Analytics
Analyze the text of users profiles
And much more
-
7/24/2019 Unit9 EHarmony AllSlides
21/25
Feature Importance: Distance
15.071x eHarmony: Maximizing the Probability of Love 6
-
7/24/2019 Unit9 EHarmony AllSlides
22/25
Feature Importance: Attractiveness
15.071x eHarmony: Maximizing the Probability of Love 7
-
7/24/2019 Unit9 EHarmony AllSlides
23/25
Feature Importance: Height Difference
15.071x eHarmony: Maximizing the Probability of Love 8
-
7/24/2019 Unit9 EHarmony AllSlides
24/25
How Successful is eHarmony?
15.071x eHarmony: Maximizing the Probability of Love 1
By 2004, eHarmony had
made over $100 million in
sales.
In 2005, 90 eHarmonymembers married every day
In 2007, 236 eHarmony
members married every day
In 2009, 542 eHarmony
members married every day
-
7/24/2019 Unit9 EHarmony AllSlides
25/25
eHarmony Maintains its Edge
15.071x eHarmony: Maximizing the Probability of Love 2
14% of the US online dating market.
The only competitor with a larger portion is
Match.com with 24%.
Nearly 4% of US marriages in2012 are a result of eHarmony.
eHarmony has successfully leveraged the power ofanalytics to create a successful and thriving business.