Data Augmentation Using Many-To-Many RNNs for Session-Aware Recommender Systems

TL;DR

This paper introduces a many-to-many RNN model for session-aware recommender systems in travel, improving efficiency and accuracy by predicting multiple destinations within user sessions.

Contribution

The paper proposes a novel many-to-many RNN approach for session-aware recommendations, enhancing computational efficiency and predictive performance over traditional methods.

Findings

Achieved 4th place in ACM WSDM WebTour 2021 Challenge

Attained an accuracy@4 of 0.5566

Demonstrated efficiency of many-to-many RNNs for session data

Abstract

The ACM WSDM WebTour 2021 Challenge organized by Booking.com focuses on applying Session-Aware recommender systems in the travel domain. Given a sequence of travel bookings in a user trip, we look to recommend the user's next destination. To handle the large dimensionality of the output's space, we propose a many-to-many RNN model, predicting the next destination chosen by the user at every sequence step as opposed to only the final one. We show how this is a computationally efficient alternative to doing data augmentation in a many-to-one RNN, where we consider every subsequence of a session starting from the first element. Our solution achieved 4th place in the final leaderboard, with an accuracy@4 of 0.5566.