1. Machine Learning & Data @ Zillow Group
Jasjeet Thind
Vice President, Data Science & Engineering
@JasjeetThind

2. Agenda Zillow Group Machine Learning Use Cases Architectural Patterns
Models
Machine Learning Pipeline
Data Quality
Free Zillow Data
@JasjeetThind

3. Zillow Group
Build the world's largest, most trusted and vibrant home-related marketplace.
@JasjeetThind

4. Machine Learning Use Cases
Personalization
Ad Targeting
Zestimate (AVMs)
Premier Agent (B2B)
Mortgages
Deep Learning
Social & Content
Demographics & Community
Business Analytics
Forecasting home price trends
@JasjeetThind

5. Real-Time Scoring APIs Data Collection Systems Collaborative Filtering
Architecture
Real-Time Scoring APIs
(Python, Flask)
Data Collection Systems
(Java/Python/SQL)
Zillow Group
Data Lake
(AWS - S3 / Kinesis)
Ranking
(Spark)
Featurization
(Spark)
User Profiles
(Spark / HBase)
Aggregate Features
(Spark)
Wedge Counting
Collaborative Filtering
(Spark)
@JasjeetThind

6. Architectural Patterns
Transport
[Collect]
Data Lake (AWS)
[Store]
Application (Backstage)
[Process]
Serving System / Analytics
[Answer]
Put object
Analytics
Put object
Get object
ZG Data Lake (S3)
Application (batch)
Kinesis
Firehose
Stream
Database (Serving)
Get object
Get records
Application (near real-time)
Real-Time Scoring
Put NRT records
@JasjeetThind

7. Machine Learning Models
K-means clustering
K-nearest neighbors
Wedge Counting
Random Forest
Gradient Boosted Machines
CNN (Deep Learning)
NLP / TF-IDF / Word2vec / Bag of Words
Linear Regression
@JasjeetThind

8. Like vs. Dislike
Predict homes per user using behavior of similar users
Like = user actively engaged with property
Dislike = user viewed property but weak engagement
Feature
Description
uid
unique id of user
pid
Property id
first_visit
timestamp or 0
num_views
sigmoid(#views)
time_spent
time on page
num_contacts
# leads sent
num_saves
# saves on zpid
num_shares
# shares on zpid
num_photos
# photos viewed
$19M + ?
Spencer
Stan - +
$22M - +
$664K
@JasjeetThind

9. Wedge Count
For all user & property pairs to form a prediction, perform wedge count
Does Stan like $19M?
Wedge # 3
(wedge03_cnt) 5
(wedge05_cnt)
$22M + -
$19M ?
Spencer
Stan
$664k - +
$19M ?
Spencer
Stan
@JasjeetThind

10. Gradient Boosting Classifier
Normalize wedge counts for popular users / properties
Prediction for all user / property pairs
features
wedge00_cnt
wedge01_cnt
wedge02_cnt
wedge03_cnt
wedge04_cnt
wedge05_cnt
wedge06_cnt
wedge07_cnt
wedge00_norm_cnt
wedge01_norm_cnt
wedge02_norm_cnt
wedge03_norm_cnt
wedge04_norm_cnt
wedge05_norm_cnt
wedge06_norm_cnt
wedge07_norm_cnt
Does Stan like the $19M home?
features
(uid: Stan, pid: $19M)
(see right side)
@JasjeetThind

11. User Profile Signals - website, mobile app, and search queries
Binary classification
labels (like/dislike) same as wedge count model
User profile model determines preference scores
Features (categorical variables)
Bath
0_bath, 0.5_bath, 1_Bath, 1.5_bath, 2_bath, 2.5_bath, 3_bath
Bed
0_bed, 1_bed, 2_bed, 3_bed, 4_bed, 5_bed
Price
100_125_price, 125_150_price, 150_175_price
Use
Code
condo, single_family, farm_land
Zipcode
zip_98109
pid
uid
features
label
0 or 1 (see right side)
0 or 1
0_bed: _bed: _bed: _bed: 0.6
@JasjeetThind

12. Ranking Property matrix feature space same as user profile
Dot product of property matrix with user profile vector
Linear regression with additional features (e.g. age decay)
0_bed
1_bed
2_bed
3_bed
uid_0
pid_0 1
0.01
0.8
0.6
0.8
0.6
pid_1
(uid, pid)
score
{"uId":" ", "pId":" "}
0.3364 =
pid_2
pid_3
@JasjeetThind

13. Machine Learning Pipeline
Collect user behavior and real-estate data, train the various models, generate the candidate set, and and make predictions.
Recommendations
Hashmap
(Redis)
Spark job creates Hive table with user events (uid, pid) partitioned by date
pid -> uid reverse index
Wedge Counting / User Profile Models
User Store
Past and current user events
Models
(Python)
User Behavior
(Kinesis /S3)
Event API
(Java)
Filter
(Spark)
Score
(Spark)
Wedge features or property features (user profile)
Public Record
(Kinesis / S3)
Training Set
(S3)
Producer
(Python)
Training Data
(Spark)
Train Models
(Spark)
Property Data
Active Listings
(Kinesis / S3)
Scoring Set
(S3)
Producer
(Python)
Scoring Data
(Spark)
Listing Data
@JasjeetThind

14. Data Quality Analytical pipelines that measure Data integrity
Attributes / outlier detection
Missing data
Expected # of records
Latency
Models - expected data
Build reports / alerts that drive action
@JasjeetThind

15. Free Zillow Data @JasjeetThind
Zillow.com/data
Zillow Home Value Index (ZHVI)
Top / Middle / Bottom Thirds
Single Family / Condo / Co-op
Median Home Value Per Sq Ft
Zillow Rent Index (ZRI)
Multi-family / SFR / Condo / Co-op
Median ZRI Per Sq ft
Median Rent List Price
Other Metrics
Median List Price
Price-to-Rent ratio
Homes Foreclosed
For-sale Inventory / Age Inventory
Negative Equity
And many more…
Time Series: national, state, metro, county, city and ZIP code levels
ZTRAX: Zillow Transaction and Assessment Dataset
Previously inaccessible or prohibitively expensive housing data for academic and institutional researchers FOR FREE.
More than 100 gigabytes
374 million detailed public records across more than 2,750 U.S. counties
20+ years of deed transfers, mortgages, foreclosures, auctions, property tax delinquencies and more for residential and commercial properties.
Assessor data including property characteristics, geographic information, and prior valuations on approximately 200 million parcels in more than 3,100 counties.
for more information
@JasjeetThind

16. Thank you! Related Blogs Zillow.com/data-science Trulia.com/blog/tech/
Hiring
Machine Learning Engineer
Data Scientist
Product Manager
Data Engineer
@JasjeetThind