KDD 19': Applying Deep Learning To Airbnb Search

🤗 Recommendation system paper challenge (27/50)

paper link

🤔 What problem do they solve?

They would like to apply deep learning model in search ranking.

😮 What are the challenges?

It is not easy to migrate to deep learning model. Usually we will try the most simple deep learning model but it is usually worser than the status quo model. In the other word, we are in plateau of Reality. However, we cannot stop there and need to keep trying a little complicated deep model.

😎 Overview of the models

They tried 4 models, and the first model failed but it did not stop their experiments of deep learning.

The second model, called Lambdarank NN, optimizing the NN for NDCG. In this way, it focused on pairwise and weighed more on first position of the rank. For example, improving the rank of a booked listing from position 2 to 1 would get priority over moving a booked listing from position 10 to 9.

The third model, called Decision Tree/Factorization Machine NN, taking the output of 3 models as input features into NN.

For more detail: Deep & Cross Network for Ad Click Predictions

The final model applying 2 hidden layers and mode LRUs on the model to avoid tech debts of previous models. In this model they try to keep everything simple except the output of these 2 model (Smart Pricing feature model) and listing embedding from co-view session.

🤔 So would they recommend deep learning to others?

That would be a wholehearted Yes.

And it’s not only because of the strong gains in the online performance of the model but also avoid too much manually feature engineering.

🥴 What else in this paper?

In this paper, they also discuss the failed model (what reason leading it failed and how do they solve it). Besides that, they talked about some detail feature and hyper parameter settings.

🙃 Other related blogs:

KDD 17': Visual Search at eBay

KDD 18': Real-time Personalization using Embeddings for Search Ranking at Airbnb

KDD 18': Notification Volume Control and Optimization System at Pinterest

KDD 19': PinText: A Multitask Text Embedding System in Pinterest

CVPR19' Complete the Look: Scene-based Complementary Product Recommendation

COLING’14: Deep Convolutional Neural Networks for Sentiment Analysis of Short Texts

NAACL’19: Utilizing BERT for Aspect-Based Sentiment Analysis via Constructing Auxiliary Sentence

NIPS’2017: Attention Is All You Need (Transformer)

KDD’19: Learning a Unified Embedding for Visual Search at Pinterest

BMVC19' Classification is a Strong Baseline for Deep Metric Learning

KDD’18: Graph Convolutional Neural Networks for Web-Scale Recommender Systems

WWW’17: Visual Discovery at Pinterest

🤩 Conference

ICCV: International Conference on Computer Vision


CVPR: Conference on Computer Vision and Pattern Recognition


KDD 2020


Top Conference Paper Challenge:


My Website:





An machine learning engineer in Bay Area in the United States

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Making virtual backgrounds by utilizing portrait segmentation

Convolutional Neural Networks: Why, what and How!

Evaluating Machine Learning Models

How to deploy NLP model on Heroku using GitHub

Creating Art with Generative Adversarial Network

Stock Price Prediction Based on Deep Learning

Image classification using Python

The Depth II: Block Matching

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Arthur Lee

Arthur Lee

An machine learning engineer in Bay Area in the United States

More from Medium

Virtual Internship Experience at LetsGrowMore

Deep Learning for Emotion Analysis

The Tooling and Dataset to capture Contextual Conversational Unsafety — why context does matter

Top 8 Machine Learning Trends for 2022