At mobile.de, Germany’s biggest car marketplace, a dedicated data team, supported by the IT project house inovex, is responsible for creating smart data products. One focus are personalised vehicle recommendations to improve the customer experience during browsing as well as finding the perfect offering. As an introduction, we briefly mention the traditional approaches for recommendation engines, thereby motivating the need for more sophisticated approaches. We then illustrate how Deep Learning can be leveraged to capture the underlying non-linear correlations of features for personalised recommendations. In particular, we’ve customised Google Play’s algorithm for an online marketplace with a fast-changing inventory. Several variants of our adapted approach are evaluated against traditional methods as well as scalability aspects are addressed. We conclude our talk by giving an outlook on the importance of personalised user experiences and the application of Deep Learning and AI at mobile.de.

1. Deep Learning-based Recommendations for
Germany’s Biggest Online Vehicle Marketplace
Bigdata.AI Summit, Hanau, March 1, 2018
Florian Wilhelm, Arnab Dutta

2. 2
Introduction
Marcel Kurovski
Data Scientist
inovex GmbH
@FlorianWilhelm
FlorianWilhelm
florianwilhelm.info
Dr. Arnab Dutta
Data Scientist
mobile.de GmbH
@kopfhohen
kraktos
o
Dr. Florian Wilhelm
Data Scientist
inovex GmbH
squall-1002

3. 3
Outline
§Introduction
§Use-cases
§Traditional
§Deep Learning
§Results

4. 4
MOBILE.DE
GERMAN MARKET
LEADER
13.5 MIO
UNIQUE USER
PER MONTH
1.6 MIO
VEHICLES
290
EMPLOYEES
DREILINDEN /
FRIEDRICHSHAIN
BERLIN
HEADQUARTERS
Part of
ebay Tech

5. 5
IT-project house for digital transformation:
‣ Agile Development & Management
‣ Web · UI/UX · Replatforming · Microservices
‣ Mobile · Apps · Smart Devices · Robotics
‣ Big Data & Business Intelligence Platforms
‣ Data Science · Data Products · Search · Deep Learning
‣ Data Center Automation · DevOps · Cloud · Hosting
‣ Trainings & Coachings
Using technology to inspire our
clients. And ourselves.
inovex offices in
Karlsruhe · Pforzheim · Köln
München · Hamburg · Stuttgart.
www.inovex.de

6. 6
Outline
§Introduction
§Use-cases
§Traditional
§Deep Learning
§Results

7. 7
Why Recommendations?Why Recommendations?
Show width of offering
Inspiration
Engagement

8. 8
- - engagement
- - inspiration
- - relevance
Why Recommendations?
- - high click-through-rate
- - small exit- & bounce-rates
User
Benefits
Business
Benefits

9. 9
X
X
X
X
Mobile.de Conversion Funnel
WishlistHome Search
Result Page
View Contact Buy
X

10. 10
Recommendations on Home
Home
Recommendations based on
preferences of visiting users as an
alternative entry point.
WishlistHome SRP View Contact Buy

11. 11
Recommendations on View Item Page
VIP
Recommendations based on the
specific make and model a user is
viewing to present alternatives
WishlistHome SRP View Contact Buy

12. 12
Recommendations on your Wishlist
Recommendations based on the
specific make and model of a
deleted ad to provide almost
identical recommendations
Recommendations based on the
users car preferences and the
parking lot items.
WishlistHome SRP View Contact Buy

13. 13
§Introduction
§Use-cases
§Traditional
§Deep Learning
§Results

14. 14
Collaborative Filtering
favourited or viewed
rated highly
also favourited or viewed
recommend
items similar

15. 15
Summary of Collaborative Filtering
üCollective behaviour of users
üStandard-Method (it works, it’s reliable etc.)
x Cold Start Problem: New listings need a
certain number of clicks to be recommended.
x Sparsity problems: lot fewer interaction
data points than total items and users.
x Content agnostic
x Only “batch-based” learning

16. 16
Looking For: Used Car (100%)
Prefers (Make): BMW (50%), Audi (50%)
Prefers (Model): Audi A3 (25%), Audi A4 (25%),
BMW 318 (50%)
Searching In: lat 52.5206, lon 13.409
Search Radius: 300km
Preferred Price: 20 000€ ± 1500€
Preferred Mileage: 10 000km ± 5000km
User Preferences
Anonymous
Content-based Filtering: User Preferences

17. 17
Content-based Filtering
interacted
<Price: 10K, Category: small>
<Price: 6K, Category: small>
<Price: 90K, Category: sports>
less similar
<Price: 10K, Category: small>
similar
recommend

18. 18
Summary of Content-based
üWorks even if there are no other
users
ücontent-based preferences of
users based on a weighted vector
of item features
xHard to do recommendations for
new users (cold start problem)
xNon-applicable for heterogenous
content types
xLow diversity, i.e. more of the same

19. 19
Traditional Hybrid Recommender
Collaborative
Filtering
Hybrid
Recommender
Content
based
PP
P P
P
Looking For: Used Car (100%)
Prefers (Make): BMW (50%), Audi (50%)
Prefers (Model): Audi A3 (25%), Audi A4 (25%),
BMW 318 (50%)
Searching In: lat 52.5206, lon 13.409
Search Radius: 300km
Preferred Price: 20 000€ ± 1500€
Preferred Mileage: 10 000km ± 5000km
User Profile
Buyer
Last Action: Yesterday
Frequent User
User 12345
Likelihood to buy: 88 %
Elastic Search Query
ü based on ES and Mahout
ü comprehensible and debuggable
ü robust and reliable concepts
ü easy to tune for different use-cases
x incapable of capturing inherent non-
linear feature dependencies
x lots of manual feature engineering

20. 20
§Introduction
§Use-cases
§Traditional
§Deep Learning
§Results

21. 21
Deep Learning
„[...] reported a 29%
sales increase to
$12.83 billion [...]“
Deep Learning Success StoriesReasons for Deep Learning
• captures nonlinear relations
• holistic approach
• less feature engineering
• improved quality
Search
Recommendations

22. 22
Find the car that perfectly fits your life
User’s Car Preferences Car Pool + Attributes
(make, model, color, price, …)
Flexible
(cold-start, uncertainty, real-time, ...)
Interactions of other users
(views, parkings, contacts)

23. 23
Recommendation Task
Deep
Black Box
Mobile.de Listings
1:
2:
4:
3:
5:
Personalized
Ranking
„Which car would the user most likely consider next?“

24. 24
Deep Learning Network
mileage
price
color
history
mileage
price
color
views
...
0.38
0.25
0.79
...
0
1
0
...
0
0.2
0.8
...
...
scalingencodescalingencode
...
0.35
-0.15
2.03
cont.
cat.
cont.
cat.
Output
Probability that user
likes vehicle
userembeddingsitemembeddings deep component
RankNet
Item
Net
User
Net

25. 25
Deep Learning Recommender - Architecture
ad storage
embeddings
RankNet
UserNet
ItemNet
Candidate
Generation
ANN Index
Candidate ServiceRanking Service
Web Service
User Preference API
Recommendation Service

26. 26
Technology Stack
Annoy ANN by
Spotify
Hardware
GPU-Server
NVIDIA Tesla K80
4x Intel Xeon 3.5 GHz
64GB RAM,
850GB Disk
LightFM
by Lyst

27. 27
§Introduction
§Motivation & Use-cases
§Theory
§Deep Learning
§Results

28. 28
Improvements by Deep Learning
0,25%
0,35%
0,45%
0,55%
0,65%
0,75%
0,85%
0,95%
1,05%
1,15%
k = 1 k = 5 k = 10 k = 30 k = 100
MAP@k
Collaborative Filtering
Traditional Hybrid
Deep Recommender
+73%
+143%

29. 29
Conclusion
Deep
Learning
Flexibility
HybridAbstraction

30. 30
Recommendations support
users to find the perfect vehicle
based on their preferences and
similarities to other users

31. 31
Any questions?lusion