• Corpus ID: 221819409

Modeling Online Behavior in Recommender Systems: The Importance of Temporal Context

  title={Modeling Online Behavior in Recommender Systems: The Importance of Temporal Context},
  author={Milena Filipovic and Blagoj Mitrevski and Diego Antognini and Emma Lejal Glaude and Boi Faltings and Claudiu Cristian Musat},
Simulating online recommender system performance is notoriously difficult and the discrepancy between the online and offline behaviors is typically not accounted for in offline evaluations. Recommender systems research tends to evaluate model performance on randomly sampled targets, yet the same systems are later used to predict user behavior sequentially from a fixed point in time. This disparity permits weaknesses to go unnoticed until the model is deployed in a production setting. We first… 

Figures and Tables from this paper

A Critical Study on Data Leakage in Recommender System Offline Evaluation
It is shown that data leakage impacts models’ recommendation accuracy, and their relative performance orders become unpredictable with different amount of future data in training, and a timeline scheme is proposed, which calls for a revisit of recommendation model design.
Recommender May Not Favor Loyal Users
A comprehensive analysis of recommendation results along global timeline finds that users who have recently interacted with the system, with respect to the time point of the test instance, enjoy better recommendations.
Exploiting Session Information in BERT-based Session-aware Sequential Recommendation
Three methods to improve recommendation performance by exploiting session information while minimizing additional parameters in a BERT-based sequential recommendation model are proposed: using session tokens, adding session segment embeddings, and a time-aware self-attention.
Report on the 1st workshop on the perspectives on the evaluation of recommender systems (PERSPECTIVES 2021) at RecSys 2021
The primary goal of the workshop was to capture the current state of evaluation from different, and maybe even diverging or contradictory perspectives.


Time-aware recommender systems: a comprehensive survey and analysis of existing evaluation protocols
A comprehensive survey and analysis of the state of the art on time-aware recommender systems (TARS), and proposes a methodological description framework aimed to make the evaluation process fair and reproducible.
Sequence-Aware Recommender Systems
A categorization of the corresponding recommendation tasks and goals is proposed, existing algorithmic solutions are summarized, methodological approaches when benchmarking what the authors call sequence-aware recommender systems are discussed, and open challenges in the area are outlined.
Multi-Gradient Descent for Multi-Objective Recommender Systems
This work introduces a stochastic multi-gradient descent approach to recommender systems (MGDRec) and shows that uncorrelated objectives, like the proportion of quality products, can be improved alongside accuracy.
Application of Dimensionality Reduction in Recommender System - A Case Study
This paper presents two different experiments where one technology called Singular Value Decomposition (SVD) is explored to reduce the dimensionality of recommender system databases and suggests that SVD has the potential to meet many of the challenges ofRecommender systems, under certain conditions.
Collaborative filtering with recency-based negative feedback
This paper proposes a recency-based scheme to perform negative preference imputation in an incremental matrix factorization algorithm designed for streaming data, and shows that this approach substantially improves the accuracy of the baseline method, outperforming both classic and state-of-the-art algorithms.
Beyond accuracy: evaluating recommender systems by coverage and serendipity
It is argued that the new ways of measuring coverage and serendipity reflect the quality impression perceived by the user in a better way than previous metrics thus leading to enhanced user satisfaction.
Recency-based collaborative filtering
This paper presents a novel algorithm namely recency-based collaborative filtering to explore the weights for items based on their expected accuracy on the future preferences, based on item- based collaborative filtering algorithms.
Incremental Collaborative Filtering recommender based on Regularized Matrix Factorization
Incorporating frequency, recency and profit in sequential pattern based recommender systems
A new sequential pattern mining algorithm that incorporates the concepts of frequency, recency, and profit to discover frequent, recent, and profitable sequential patterns, called FRP-sequences is proposed.
Optimizing the Recency-Relevancy Trade-off in Online News Recommendations
This work proposes a new recommendation strategy (called Highest Future-Impact) which attempts to optimize on both the axes of recency and relevancy, and develops an optimization framework combining the predicted future-impact of the stories with the uncertainties in the predictions.