Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up...
Transcript of Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up...
![Page 1: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/1.jpg)
Which car fits my life?Mobile.de’s approach to recommendations
PyData Berlin, July 1st, 2017Florian Wilhelm, Arnab Dutta
![Page 2: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/2.jpg)
2
Introduction
Dr. Florian Wilhelm
Data Scientist
inovex GmbH
� @FlorianWilhelm
� FlorianWilhelm
florianwilhelm.info
Dr. Arnab Dutta
Data Scientist
mobile.de GmbH
� @kopfhohen
� kraktoso
![Page 3: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/3.jpg)
3
Outline
§Introduction§Use-cases§Theory§Our Approach§Example§Outlook
![Page 4: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/4.jpg)
4
MOBILE.DEGERMAN MARKET LEADER
13.5 MIO UNIQUE USER PER MONTH
1.6 MIO VEHICLES
290EMPLOYEES
DREILINDEN / FRIEDRICHSHAIN BERLINHEADQUARTERS
Part ofebay Tech
![Page 5: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/5.jpg)
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 inKarlsruhe · Pforzheim · München · Köln · Hamburg.
www.inovex.de
![Page 6: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/6.jpg)
6
Outline
§Introduction§Use-cases§Theory§Our Approach§Example§Outlook
![Page 7: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/7.jpg)
7
Why Recommendations?Why Recommendations?
Show width of offering
Inspiration
Engagement
![Page 8: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/8.jpg)
8
- - engagement- - inspiration- - relevance
Why Recommendations?
- - high click-through-rate - - small exit- & bounce-rates
User Benefits
Business Benefits
![Page 9: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/9.jpg)
9
Mobile.de Conversion Funnel
WishlistHome Search Result Page
View Contact Buy
![Page 10: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/10.jpg)
10
Recommendations on Home
Home
Recommendations based on preferencesof visiting users as an alternative entry point.
WishlistHome SRP View Contact Buy
![Page 11: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/11.jpg)
11
Recommendations on Search Results Page
Recommendations basedon similar vehicle makeand model id to presentalternatives
WishlistHome SRP View Contact Buy
![Page 12: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/12.jpg)
12
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
![Page 13: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/13.jpg)
13
Recommendations on your Wishlist
Recommendations based on the specific make and model of a deleted ad to provide almost identical recommendations
Recommendations based on theusers car preferences and the parking lot items.
WishlistHome SRP View Contact Buy
![Page 14: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/14.jpg)
14
§Introduction§Use-cases§Theory§Collaborative Filtering§Content Based§Our Approach§Example§Outlook
![Page 15: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/15.jpg)
15
Collaborative Filtering
items similar
![Page 16: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/16.jpg)
16
Watched / Rated
Unwatched
Item-Item Similarity
??
??
????Item-based Recommendations
Cosine Similarity
![Page 17: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/17.jpg)
17
Recommendation
Item-based Recommendations
P
P
P
P
P P
Wishlist
![Page 18: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/18.jpg)
18
Summary of Collaborative Filtering✓Collective behaviour of users
✓Standard-Method (it works, it’s reliable etc.)
� Cold Start Problem: New listings need a certain number of clicks to be recommended.
� Sparsity problems: lot fewer interactiondata points than total items and users.
� Content agnostic
� Only “batch-based” learning
![Page 19: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/19.jpg)
19
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 13.409Search Radius: 300kmPreferred Price: 20 000€ ± 1500€ Preferred Mileage: 10 000km ± 5000km
User Preferences
Marketing
Anonymous
Content-based Filtering: User Preferences
![Page 20: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/20.jpg)
20
Content-based Filtering
interacted
<Price: 10K, Category: small>
<Price: 6K, Category: small>
<Price: 90K, Category: sports>
<Price: 10K, Category: small>
recommend
![Page 21: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/21.jpg)
21
Summary of Content-basedüWorks even if there are no other
users
ücontent-based preferences ofusers based on a weighted vectorof item features
� Hard to do recommendations fornew users (cold start problem)
� Non-applicable for heterogenouscontent types
� Low diversity, i.e. more of thesame
![Page 22: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/22.jpg)
22
§Introduction§Use-cases§Theory§Our Approach§Example§Outlook
![Page 23: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/23.jpg)
23
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)
![Page 24: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/24.jpg)
24
Hybrid Recommender
Collaborative Filtering
Hybrid Recommender
Contentbased
• no cold-start problem of new items
• integrate new user events in real-time
• robust and reliable concepts
• easy to tune for different use-cases
• comprehensible and debuggable
![Page 25: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/25.jpg)
25
RecommendationEngine
User PreferenceService API
User EventTracking & Storage
RecommendationEngine
RecommendationEngine
User PreferenceComputation &
Storage
All Listings
User Preference+ Recommendation Architecture
![Page 26: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/26.jpg)
26
Hybrid Recommender Concept
PPP 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 13.409Search Radius: 300kmPreferred Price: 20 000€ ± 1500€Preferred Mileage: 10 000km ± 5000km
User Profile
BuyerLast Action: YesterdayFrequent User
User 12345
Likelihood to buy: 88 %
Elastic Search Query
Score0.8
0.3
1.7
3.2
1.1
0.9
……
![Page 27: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/27.jpg)
27
§Introduction§Use-cases§Theory§Our Approach§Example§Outlook
![Page 28: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/28.jpg)
28
Finding similar make/models
§Users are often uncertainwith their choices in orientation phase
§Help users explore similarmodels to make informeddecisions
§Exploit the collaborativeaspect in defining theconcept of similarity
![Page 29: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/29.jpg)
29
Make/Model Recommender with LightFM
LightFM:§Matured and well documented Python package§Optimized and parallelized with Cython§Hybrid recommender based on matrix factorisation§Supports Learning-to-Rank objectives (BPR, WARP)
Audi A4 BMW 645 similar vehicles
![Page 30: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/30.jpg)
30
Non-negative Matrix Factorisation (NMF)LF1 LF2
LF1
LF2
M (|U| x |I|) x R (|LF| x |I|)
= X
= L (|U| x |LF|)
![Page 31: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/31.jpg)
31
NMF: Embeddings to Similarities
LF1
LF2
• car is represented as an item embedding
• Given 2 embeddings, LFitem_i and Lfitem_j_
• Compute sim(LFitem_i , LFitem_j)
• Find pairwise values for all item pairs (|I| x |I|)
item embeddings
![Page 32: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/32.jpg)
32
NMF: Persisting Item similarities
![Page 33: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/33.jpg)
33
Method§ best of both the worlds§ robust
Business§ higher CTR§ lesser exits rates
User§ engagement§ diversity
hybrid
![Page 34: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/34.jpg)
34
§Introduction§Motivation & Use-cases§Theory§Implementation§Outlook
![Page 35: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/35.jpg)
35
Deep Learning
Dermatologist-level classification of skin cancer with deep neural networks(Nature 542, 115-118, February 2017)
DeepStack: Expert-level artificial intelligence in heads-up no-limit poker(Science, March 2017)
Recent Breakthroughs in Deep Learning Reasons for Deep Learning
• captures nonlinear relations
• holistic approach, i.e. reduces number of components possibly
• less feature engineering
• possibly improved quality
![Page 36: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/36.jpg)
36
Approach: Wide and Deep Model
mileagepricecolor
history
mileagepricecolor
views
...
0.38
0.250.79
...0
10
...0
0.20.8
...
...
enco
deen
code
enco
deen
code
...
0.35
-0.152.03
cont.
cat.
cont.
cat.us
erem
bedd
ings
item
em
bedd
ings
Outp
ut
cros
s-ite
m
Probability that user X likes vehicle Y
Deep Component
Wide Component
one-
hot
![Page 37: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/37.jpg)
37
Recommendations support users to find the perfect vehicle based on their preferences and by collaboration
![Page 38: Which car fits my life? - inovex · DeepStack: Expert-level artificial intelligence in heads-up no-limit poker (Science, March 2017) Recent Breakthroughs in Deep Learning Reasons](https://reader034.fdocumento.com/reader034/viewer/2022051606/60172098e337e759c53d102d/html5/thumbnails/38.jpg)
38
Any questions?lusion