Parallel Recurrent Neural Network Architectures for Feature-rich Session-based Recommendations

By Balázs Hidasi, Massimo Quadrana, Alexandros Karatzoglou, Domonkos Tikk. 2016

544 citations recorded by the source corpus.

Real-life recommender systems often face the daunting task of providing recommendations based only on the clicks of a user session. Methods that rely on user profiles -- such as matrix factorization -- perform very poorly in this setting, thus item-to-item recommendations are used most of the time. However the items typically have rich feature representations such as pictures and text descriptions that can be used to model the sessions. Here we investigate how these features can be exploited in Recurrent Neural Network based session models using deep learning. We show that obvious approaches do not leverage these data sources. We thus introduce a number of parallel RNN (p-RNN) architectures to model sessions based on the clicks and the features (images and text) of the clicked items. We also propose alternative training strategies for p-RNNs that suit them better than standard training. We show that p-RNN architectures with proper training have significant performance improvements over feature-less session models while all session-based models outperform the item-to-item type baseline.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Information Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Information Systems

Citations in the corpus

Contextual Sequence Modeling for Recommendation with Recurrent Neural Networks (2017, background): “In a follow-up paper [6], Hidasi et al.”
Recurrent Neural Networks with Top-k Gains for Session-based Recommendations (2017, positive): “Since the size of the VIDXL test set is large, we compare the target item’s score to that of the 50,000 most popular items during testing, similarly to [8].”
Deep Learning Based Recommender System (2017, background): “Two extensions [57, 133] demonstrate that side information has an effect on enhancing session recommendation quality.”
STAMP: Short-Term Attention/Memory Priority Model for Session-based Recommendation (2018, background): “Inspired by recent advances in natural language processing area [16], some deep learning based solutions have been developed and some of which represent the state-of-the-art in SRS research field [2, 5, 6, 10].”
Personalizing Session-based Recommendations with Hierarchical Recurrent Neural Networks (2017, methodology): “To speed up evaluation, we computed the rank of the relevant item compared to the 50,000 most supported items, as done in [8].”
A Unified Deep Learning Architecture for Abuse Detection (2018, methodology): “systems [19], and explain all technical details needed for its repeatability.”
When Recurrent Neural Networks meet the Neighborhood for Session-Based Recommendation (2017, background): “In [15] and [32] RNN-based approaches were recently proposed that leverage additional item features to achieve higher accuracy.”
A Survey on Accuracy-Oriented Neural Recommendation: From Collaborative Filtering to Information-Rich Recommendation (2021, methodology): “do not log in and user IDs are not available [201], [202], [203].”
Filter-enhanced MLP is All You Need for Sequential Recommendation (2022, methodology): “A series of works follow this line and extend it to high-order MCs [18, 22, 53].”
Causal embeddings for recommendation (2017, background): “We can separate the newly proposed solutions in two main types of approaches, ones that learn item embeddings that optimize for item-item similarity prediction ([7, 16, 24]) and the ones that learn user sequence embeddings to optimize for next item prediction ([8, 9]).”
Evaluation of session-based recommendation algorithms (2018, methodology): “…modeling approaches are usually based on Markov Chain (MC) models (He et al. 2009; Mcfee and Lanckriet 2011; Garcin et al. 2013; Hosseinzadeh Aghdam et al. 2015), reinforcement learning (RL) andMarkov Decision Processes (MDP) (Shani et al. 2005;Moling et al. 2012; Tavakol and Brefeld 2014),…”
Improving Sequential Recommendation with Knowledge-Enhanced Memory Networks (2018, methodology): “[23] propose to incorporate auxiliary features into GRU networks for improving the sequential recommendation.”
Short-Term Satisfaction and Long-Term Coverage: Understanding How Users Tolerate Algorithmic Exploration (2018, methodology): “As an intermediate step, future work could combine interactive recommendation with techniques for session-based recommendation [19, 21, 43].”
Deep Session Interest Network for Click-Through Rate Prediction (2019, methodology): “General Factorization Framework (GFF) [ Hidasi and Tikk, 2016 ] uses sum pooling of items to represent a session.”
A review on deep learning for recommender systems: challenges and remedies (2018, methodology): “RNNs are utilized in recommender systems to integrate current viewing web page history and order of the views to provide more accurate recommendations (Wu et al. 2016a; Hidasi et al. 2016a; Tan et al. 2016).”
Session-based Recommendation with Graph Neural Networks (2018, methodology): “For session-based recommendation, the work of (Hidasi et al. 2016a) proposes the recurrent neural network approach, and then extends to an architecture with parallel RNNs (Hidasi et al. 2016b) which can model sessions based on the clicks and features of the clicked items.”
Explainable Interaction-driven User Modeling over Knowledge Graph for Sequential Recommendation (2019, methodology): “Sequence modeling methods for SR mainly belong to Markov Models [4, 17] and RNNs[8, 9, 15, 27].”
Sequential Recommender Systems: Challenges, Progress and Prospects (2019, background): “On the one hand, to take the influences of items’ features on the user-item interactions into account, a series of feature-enriched neural models including [9] have been proposed for SRSs.”