The Simpson’s Paradox in the Offline Evaluation of Recommendation Systems
TL;DR: In this situation, users only provide feedback that were collected from an existing, already deployed recommendation system as discussed by the authors, and users only provided feedback was used to evaluate the recommendation system based on user interactions.
read more
Abstract: Recommendation systems are often evaluated based on user’s interactions that were collected from an existing, already deployed recommendation system. In this situation, users only provide feedback ...
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Agent-Based Recommendation in E-Learning Environment Using Knowledge Discovery and Machine Learning Approaches
Zeinab Shahbazi,Yung-Cheol Byun +1 more
TL;DR: This research proposes virtual and intelligent agent-based recommendation, which requires users’ profile information and preferences to recommend the proper content and search results based on their search history, and applies Natural Language Processing techniques and semantic analysis approaches for the recommendation of course selection to e-learners and tutors.
Off-Policy Actor-critic for Recommender Systems
Minmin Chen,Can Xu,Vince Gatto,Devanshu Jain,Aviral Kumar,Ed H. Chi +5 more
- 18 Sep 2022
TL;DR: The key designs in setting up an off-policy actor-critic agent for production recommender systems are shared and it is demonstrated in offline and live experiments that the new framework out-performs baseline and improves long term user experience.
OCA: Ordered Clustering-Based Algorithm for E-Commerce Recommendation System
TL;DR: The Ordered Clustering-based Algorithm (OCA) as discussed by the authors attempts to exploit the collaborative filtering strategy for e-commerce recommendation systems to cluster users based on their similarities in preferences.
A Critical Study on Data Leakage in Recommender System Offline Evaluation
TL;DR: In this article , the authors provide a comprehensive and critical analysis of the data leakage issue in recommender system offline evaluation, e.g., train/test data split does not follow global timeline.
•Posted Content
Causal Collaborative Filtering.
TL;DR: In this article, a general framework for modeling causality in collaborative filtering and recommendation is proposed, which is based on a unified causal view of CF and mathematically shows that many traditional CF algorithms are actually special cases of CCF under simplified causal graphs.
20
References
Matrix Factorization Techniques for Recommender Systems
TL;DR: As the Netflix Prize competition has demonstrated, matrix factorization models are superior to classic nearest neighbor techniques for producing product recommendations, allowing the incorporation of additional information such as implicit feedback, temporal effects, and confidence levels.
An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies
TL;DR: The propensity score is a balancing score: conditional on the propensity score, the distribution of observed baseline covariates will be similar between treated and untreated subjects, and different causal average treatment effects and their relationship with propensity score analyses are described.
10.3K
Neural Collaborative Filtering
Xiangnan He,Lizi Liao,Hanwang Zhang,Liqiang Nie,Xia Hu,Tat-Seng Chua +5 more
- 03 Apr 2017
TL;DR: This work strives to develop techniques based on neural networks to tackle the key problem in recommendation --- collaborative filtering --- on the basis of implicit feedback, and presents a general framework named NCF, short for Neural network-based Collaborative Filtering.
Evaluating collaborative filtering recommender systems
TL;DR: The key decisions in evaluating collaborative filtering recommender systems are reviewed: the user tasks being evaluated, the types of analysis and datasets being used, the ways in which prediction quality is measured, the evaluation of prediction attributes other than quality, and the user-based evaluation of the system as a whole.
The MovieLens Datasets: History and Context
TL;DR: The history of MovieLens and the MovieLens datasets is documents, including a discussion of lessons learned from running a long-standing, live research platform from the perspective of a research organization, and best practices and limitations of using the Movie Lens datasets in new research are documented.
5K