Deep Learning-based Recommendations for Germany’s Biggest Online Vehicle Marketplace – mobile.de

Empfehlungen basierend auf Deep-Learning-Ansätzen für Deutschlands größten Online-Fahrzeugmarkt.

Abstract:

At mobile.de, Germany’s biggest car marketplace, a dedicated team of data engineers and scientists, supported by the IT project house inovex is responsible for creating intelligent data products. Driven by our company slogan “Find the car that fits your life”, we focus on personalized recommendations to address several user needs. Thereby we improve customer experience during browsing as well as finding the perfect offering.

In an introduction to recommendation systems, we briefly mention the traditional approaches for recommendation engines, thereby motivating the need for more sophisticated approaches. In particular, we explain the different concepts like collaborative and content-based filtering as well as hybrid approaches and general matrix factorization methods.

As a highly promising methodology for recommendation engines, we illustrate how Deep Learning can be leveraged in order to capture the underlying non-linear correlations of features for personalized recommendations in an end-to-end approach. In particular, we elaborate on an adaptation as well as improvements of Google Play’s recommendation engine for the special requirements of an online marketplace with a fast-changing inventory like mobile.de. This is accomplished by augmentation of the optimization criterion with embedding similarity besides the user’s preferences. Several variants of our adapted Deep Learning approach are evaluated and compared with each other as well as with a baseline given by recommendation engine based on more traditional methods.

As another important requirement of modern recommendation engine, the challenge of scaling to a large number of requests is addressed with the help of the learning-torank dichotomy which divides the task into the generation of candidates and their ranking.

We conclude our talk by giving an outlook on the importance of personalized user experiences and the application of Deep Learning & AI at mobile.de.

Speaker: Dr. Florian Wilhelm (inovex), Dr. Arnab Dutta (mobile.de)

Event: AI Summit 2017

Datum: 01.03.2018

Dr. Florian Wilhelm

Dr. Florian Wilhelm ist Data Scientist bei inovex. Er verfügt über mehrjährige Projekterfahrung im Bereich Predictive & Prescriptive Analytics und Big Data und über fundierte Kenntnisse in den Bereichen mathematische Modellierung, Statistik, maschinelles Lernen, Hochleistungsrechnen und Data Mining.

Sie haben Fragen oder möchten zu diesem Thema beraten werden?

Dann rufen Sie uns an unter +49 (0)721 619 021-0 oder schreiben Sie uns eine E-Mail. Wir freuen uns auf Ihre Nachricht!

inovex Events

Hier geht's zu unseren aktuellen Messen, Konferenzen und Meetups.

Zur Event-Liste

inovex Blog

Florian Wilhelm I 17.01.2018

Data Science in Production: Packaging, Versioning and Continuous Integration

Blog-Artikel lesen

Aktuelle News über inovex und die Branche

inovex Newsletter

Jetzt hier anmelden!

Arbeiten bei inovex

inovexperts erzählen ...

Was es konkret im Alltag bedeutet, bei inovex als „inovexpert“ zu arbeiten, erklären die Kolleginnen und Kollegen in unseren inovexpert-Videos. Wir wünschen viel Spaß beim Ansehen!

Jetzt inovexperts-Videos ansehen

inovex Case Study

mobile.de: Use Cases für Recommendations bei Online-Portalen anhand von Datenprodukten

Case Study lesen