Neural Networks for Information Retrieval

Machine learning plays an important role in many aspects of modern IR systems, and deep learning is applied to all of those. The fast pace of modern-day research into deep learning has given rise to many different approaches to many different IR problems.

  Download slides (PDF, 30MB)

Full-day tutorial at WSDM 2018 on February 5, 2018

Table of contents

The tutorial is organized in 10 parts as described below. For more details, see the overview paper on arXiv.

Time slot Topic Presenters
09:00 - 09:05 Introduction Alexey Borisov section slides
09:05 - 09:45 Preliminaries Alexey Borisov section slides
09:45 - 10:30 Modeling User Behavior Alexey Borisov section slides
10:30 - 11:00 Coffee break
11:00 - 11:45 Semantic matching Christophe Van Gysel section slides
11:45 - 12:30 Learning to rank Christophe Van Gysel, Hosein Azarbonyad section slides
12:30 - 14:00 Lunch break
14:00 - 14:45 Entities Tom Kenter, Christophe Van Gysel section slides
14:45 - 15:30 Generating Responses Tom Kenter section slides
15:00 - 15:30 Coffee break
15:30 - 16:00 Generating Responses (Part II) Tom Kenter section slides
16:00 - 16:45 Recommender systems Hosein Azarbonyad section slides
16:45 - 17:00 Industry insights Tom Kenter section slides
17:00 - 17:30 Q + A Tom Kenter, Christophe Van Gysel, Hosein Azarbonyad

ReferencesAll slides


If you wish to refer to the tutorial in your scientific publication, please refer to our overview paper:

  Author = {Kenter, Tom and Borisov, Alexey and Van Gysel, Christophe and Dehghani, Mostafa and de Rijke, Maarten and Mitra, Bhaskar},
  Booktitle = {WSDM 2018},
  Organization = {ACM},
  Title = {Neural Networks for Information Retrieval},
  Year = {2018}}