Recommendations as treatments: Debiasing learning and evaluation |
Methods for Selection Bias |
Debiasing in evaluation |
|
MF-IPS |
2016 |
PMLR |
Pyhton |
Training and testing of recommender systems on data missing not at random |
Methods for Selection Bias |
Debiasing in evaluation |
|
|
2010 |
KDD |
|
Probabilistic matrix factorization with non-random missing data |
Methods for Selection Bias |
Debiasing in model training |
|
MF-MNAR |
2014 |
JMLR |
Python |
Collaborative filtering and the missing at random assumption |
Methods for Selection Bias |
Debiasing in model training |
|
|
2007 |
UAI |
Python |
Collaborative prediction and ranking with non-random missing data |
Methods for Selection Bias |
Debiasing in model training |
|
MM |
2009 |
RecSys |
|
Social recommendation with missing not at random data |
Methods for Selection Bias |
Debiasing in model training |
|
SPMF-MNAR |
2018 |
ICDM |
|
The deconfounded recommender: A causal inference approach to recommendation |
Methods for Selection Bias |
Debiasing in model training |
|
Deconfounded Probabilistic / Poisson/Weighted MF |
2018 |
arXiv |
|
Evaluation of recommendations: rating-prediction and ranking |
Methods for Selection Bias |
Debiasing in model training |
|
|
2013 |
RecSys |
|
Training and testing of recommender systems on data missing not at random |
Methods for Selection Bias |
Debiasing in model training |
|
|
2010 |
KDD |
|
Recommendations as treatments: Debiasing learning and evaluation |
Methods for Selection Bias |
Debiasing in model training |
|
MF-IPS |
2020 |
SIGIR |
Python |
Doubly robust joint learning for recommendation on data missing not at random |
Methods for Selection Bias |
Debiasing in model training |
|
DR |
2019 |
ICML |
|
Asymmetric tri-training for debiasing missing-not-at-random explicit feedback |
Methods for Selection Bias |
Debiasing in model training |
|
AT |
2020 |
SIGIR |
|
Are you influenced by others when rating?: Improve rating prediction by conformity modeling |
Methods for Conformity Bias |
|
|
|
2016 |
RecSys |
|
Xgboost: A scalable tree boosting system |
Methods for Conformity Bias |
|
|
XGBoost |
2016 |
KDD |
Python |
Learning to recommend with social trust ensemble |
Methods for Conformity Bias |
|
|
RSTE |
2009 |
SIGIR |
|
mtrust: discerning multi-faceted trust in a connected world |
Methods for Conformity Bias |
|
|
mTrust |
2012 |
WSDM |
|
A probabilistic model for using social networks in personalized item recommendation |
Methods for Conformity Bias |
|
|
SPF |
2015 |
RecSys |
Python |
Learning personalized preference of strong and weak ties for social recommendation |
Methods for Conformity Bias |
|
|
PTPMF |
2017 |
WWW |
|
Unbiased offline recommender evaluation for missing-not-at-random implicit feedback |
Methods for Exposure Bias |
Debiasing in evaluation |
|
|
2018 |
RecSys |
Python |
Collaborative filtering for implicit feedback datasets |
Methods for Exposure Bias |
Debiasing in model training |
|
|
2008 |
ICDM |
Python |
Dynamic matrix factorization with priors on unknown values |
Methods for Exposure Bias |
Debiasing in model training |
|
|
2015 |
KDD |
C++ |
Mind the gaps: weighting the unknown in large-scale one-class collaborative filtering |
Methods for Exposure Bias |
Debiasing in model training |
|
GALS/sMMMF-ENS |
2009 |
KDD |
|
One-class collaborative filtering |
Methods for Exposure Bias |
Debiasing in model training |
|
wALS/sALS-ENS |
2008 |
ICDM |
|
Fast matrix factorization for online recommendation with implicit feedback |
Methods for Exposure Bias |
Debiasing in model training |
|
eALS |
2016 |
SIGIR |
Python |
Selection of negative samples for one-class matrix factorization |
Methods for Exposure Bias |
Debiasing in model training |
|
Full |
2017 |
SDM |
Python |
Improving one-class collaborative filtering by incorporating rich user information |
Methods for Exposure Bias |
Debiasing in model training |
|
|
2010 |
CIKM |
|
Unbiased recommender learning from missing-not-at-random implicit feedback |
Methods for Exposure Bias |
Debiasing in model training |
|
Rel-MF |
2020 |
WSDM |
Python |
Logistic matrix factorization for implicit feedback data |
Methods for Exposure Bias |
Debiasing in model training |
|
LogisticMF |
2014 |
NIPS |
|
Bpr: Bayesian personalized ranking from implicit feedback |
Methods for Exposure Bias |
Debiasing in model training |
|
BPR |
2009 |
UAI |
Python |
Collaborative denoising auto-encoders for top-n recommender systems (CDAE) |
Methods for Exposure Bias |
Debiasing in model training |
|
CDAE |
2016 |
WSDM |
C++ |
Neural collaborative filtering (NCF) |
Methods for Exposure Bias |
Debiasing in model training |
|
NCF |
2017 |
WWW |
Python |
Lightgcn: Simplifying and powering graph convolution network for recommendation (LightGCN) |
Methods for Exposure Bias |
Debiasing in model training |
|
LightGCN |
2020 |
SIGIR |
Python |
Selection of negative samples for one-class matrix factorization |
Methods for Exposure Bias |
Debiasing in model training |
|
Full |
2017 |
SDM |
Python |
Reinforced negative sampling for recommendation with exposure data |
Methods for Exposure Bias |
Debiasing in model training |
|
RNS |
2019 |
IJCAI |
Python |
An improved sampler for bayesian personalized ranking by leveraging view data |
Methods for Exposure Bias |
Debiasing in model training |
|
BPR+view |
2018 |
WWW |
|
Samwalker: Social recommendation with informative sampling strategy |
Methods for Exposure Bias |
Debiasing in model training |
|
Samwalker |
2019 |
WWW |
MATLAB/C++ |
Reinforced negative sampling over knowledge graph for recommendation |
Methods for Exposure Bias |
Debiasing in model training |
|
KGPolicy |
2020 |
WWW |
Python |
“Modeling users’ exposure with social knowledge influence and consumption influence for recommendation |
Methods for Exposure Bias |
Debiasing in model training |
|
SoEXBMF |
2018 |
CIKM |
|
Collaborative filtering with social exposure: A modular approach to social recommendation |
Methods for Exposure Bias |
Debiasing in model training |
|
SERec |
2018 |
AAAI |
C++ |
Modeling user exposure in recommendation |
Methods for Exposure Bias |
Debiasing in model training |
|
Content ExpoMF/Location ExpoMF |
2016 |
WWW |
Python |
Samwalker: Social recommendation with informative sampling strategy |
Methods for Exposure Bias |
Debiasing in model training |
|
Samwalker |
2019 |
WWW |
MATLAB/C++ |
Fast adaptively weighted matrix factorization for recommendation with implicit feedback |
Methods for Exposure Bias |
Debiasing in model training |
|
FAWMF |
2020 |
AAAI |
|
Learning to rank with selection bias in personal search |
Methods for Exposure Bias |
Debiasing in model training |
|
Global/Segmented/ Generalized Bias Model |
2016 |
SIGIR |
|
Correcting for selection bias in learning-to-rank systems |
Methods for Exposure Bias |
Debiasing in model training |
|
Heckman rank/Propensity SVM rank |
2020 |
WWW |
Python |
Large-scale causal approaches to debiasing post-click conversion rate estimation with multi-task learning |
Methods for Exposure Bias |
Debiasing in model training |
|
Multi-IPW/Multi-DR |
2020 |
WWW |
|
Entire space multi-task model: An effective approach for estimating post-click conversion rate |
Methods for Exposure Bias |
Debiasing in model training |
|
ESMM |
2018 |
SIGIR |
|
Entire space multi-task modeling via post-click behavior decomposition for conversion rate prediction |
Methods for Exposure Bias |
Debiasing in model training |
|
ESM^2^ |
2020 |
SIGIR |
|
Gmcm: Graph-based micro-behavior conversion model for post-click conversion rate estimation |
Methods for Exposure Bias |
Debiasing in model training |
|
GMCM |
2020 |
SIGIR |
|
” click” is not equal to” like”: Counterfactual recommendation for mitigating clickbait issue |
Methods for Exposure Bias |
Debiasing in model training |
|
CR |
2020 |
arXiv |
|
An experimental comparison of click position-bias models |
Methods for Position Bias |
Click models |
|
|
2008 |
WSDM |
|
A user browsing model to predict search engine click data from past observations. |
Methods for Position Bias |
Click models |
|
|
2008 |
SIGIR |
|
A dynamic bayesian network click model for web search ranking |
Methods for Position Bias |
Click models |
|
DBN |
2009 |
WWW |
|
Comparing click logs and editorial labels for training query rewriting |
Methods for Position Bias |
Click models |
|
|
2007 |
WWW |
|
Click chain model in web search |
Methods for Position Bias |
Click models |
|
CCM |
2009 |
WWW |
|
A novel click model and its applications to online advertising |
Methods for Position Bias |
Click models |
|
GCM |
2010 |
WSDM |
|
A deep recurrent survival model for unbiased ranking |
Methods for Position Bias |
Click models |
|
DRSR |
2020 |
SIGIR |
Python |
Unbiased learning-to-rank with biased feedback |
Methods for Position Bias |
Propensity score |
|
Propensity SVM-Rank |
2017 |
WSDM |
Link |
Learning to rank with selection bias in personal search |
Methods for Position Bias |
Propensity score |
|
Global/Segmented/ Generalized Bias Model |
2016 |
SIGIR |
|
Multileave gradient descent for fast online learning to rank |
Methods for Position Bias |
Propensity score |
|
|
2016 |
WSDM |
|
Learning socially optimal information systems from egoistic users |
Methods for Position Bias |
Propensity score |
|
SoPer-R/SoPer-S |
2013 |
ECML PKDD 2013 |
|
Batch learning from logged bandit feedback through counterfactual risk minimization |
Methods for Position Bias |
Propensity score |
|
POEM |
2015 |
JMLR |
|
Reusing historical interaction data for faster online learning to rank for ir |
Methods for Position Bias |
Propensity score |
|
RHC/CPS |
2013 |
WSDM |
Link |
Unbiased learning to rank with unbiased propensity estimation |
Methods for Position Bias |
Propensity score |
|
DLA |
2018 |
SIGIR |
Python |
Position bias estimation for unbiased learning to rank in personal search |
Methods for Position Bias |
Propensity score |
|
Regression-based EM |
2018 |
WSDM |
|
Attribute-based propensity for unbiased learning in recommender systems: Algorithm and case studies |
Methods for Position Bias |
Propensity score |
|
|
2020 |
KDD |
|
Debiasing grid-based product search in e-commerce |
Methods for Position Bias |
Propensity score |
|
|
2020 |
KDD |
|
Cascade model-based propensity estimation for counterfactual learning to rank |
Methods for Position Bias |
Propensity score |
|
CM-IPS |
2020 |
SIGIR |
Python |
Controlling popularity bias in learning-to-rank recommendation |
Methods for Popularity Bias |
Regularization |
|
|
2017 |
RecSys |
|
Incorporating diversity in a learning to rank recommender system |
Methods for Popularity Bias |
Regularization |
|
|
2016 |
FLAIRS |
|
Correcting popularity bias by enhancing recommendation neutrality |
Methods for Popularity Bias |
Regularization |
|
|
2014 |
RecSys |
|
Efficiency improvement of neutrality-enhanced recommendation |
Methods for Popularity Bias |
Regularization |
|
|
2013 |
RecSys |
Link |
ESAM: discriminative domain adaptation with non-displayed items to improve long-tail performance |
Methods for Popularity Bias |
Regularization |
|
ESAM |
2020 |
SIGIR |
Python |
An adversarial approach to improve long-tail performance in neural collaborative filtering |
Methods for Popularity Bias |
Adversarial learning |
|
|
2018 |
CIKM |
|
Disentangling User Interest and Conformity for Recommendation with Causal Embedding |
Methods for Popularity Bias |
Causal graph |
|
DICE |
2021 |
WWW |
Python |
The limits of popularity-based recommendations, and the role of social ties |
Methods for Popularity Bias |
Others methods |
|
|
2016 |
KDD |
C++ |
Popularity bias in ranking and recommendation |
Methods for Popularity Bias |
Others methods |
|
|
2019 |
AIES |
|
Unbiased offline recommender evaluation for missing-not-at-random implicit feedback |
Methods for Popularity Bias |
Others methods |
|
|
2018 |
RecSys |
Python |
Discrimination-aware data mining |
Methods for Unfairness |
Rebalancing |
|
|
2008 |
KDD |
|
Fairness-aware ranking in search & recommendation systems with application to linkedin talent search |
Methods for Unfairness |
Rebalancing |
|
PSL |
2019 |
KDD |
|
Designing fair ranking schemes |
Methods for Unfairness |
Rebalancing |
|
|
2019 |
SIGMOD |
|
Equity of attention: Amortizing individual fairness in rankings |
Methods for Unfairness |
Rebalancing |
|
|
2018 |
SIGIR |
|
Fa*ir: A fair top-k ranking algorithm |
Methods for Unfairness |
Rebalancing |
|
FA*IR |
2017 |
CIKM |
Java/Python |
Fairness of exposure in rankings |
Methods for Unfairness |
Rebalancing |
|
|
2018 |
KDD |
|
A fairness-aware hybrid recommender system |
Methods for Unfairness |
Rebalancing |
|
PSL |
2018 |
FATREC |
|
Fairwalk: Towards fair graph embedding |
Methods for Unfairness |
Rebalancing |
|
Fairwalk |
2019 |
IJCAI |
|
Debayes: a bayesian method for debiasing network embeddings |
Methods for Unfairness |
Rebalancing |
|
DeBayes |
2020 |
ICML |
Link |
Learning fair representations |
Methods for Unfairness |
Regularization |
|
|
2013 |
JMLR |
|
Enhancement of the neutrality in recommendation |
Methods for Unfairness |
Regularization |
|
|
2012 |
RecSys |
|
Efficiency improvement of neutrality-enhanced recommendation. |
Methods for Unfairness |
Regularization |
|
|
2013 |
RecSys |
Link |
Model-based approaches for independence-enhanced recommendation |
Methods for Unfairness |
Regularization |
|
|
2016 |
IEEE |
Link |
Considerations on recommendation independence for a find-good-items task |
Methods for Unfairness |
Regularization |
|
IERS |
2017 |
Workshop on Responsible Recommendation |
|
Beyond parity: Fairness objectives for collaborative filtering |
Methods for Unfairness |
Regularization |
|
|
2017 |
NIPS |
|
New fairness metrics for recommendation that embrace differences |
Methods for Unfairness |
Regularization |
|
|
2017 |
FAT/ML |
|
Fairness-aware group recommendation with pareto-efficiency |
Methods for Unfairness |
Regularization |
|
|
2017 |
RecSys |
|
Balanced neighborhoods for fairness-aware collaborative recommendation |
Methods for Unfairness |
Regularization |
|
SLIM |
2017 |
FATRec |
|
Fairness in recommendation ranking through pairwise comparisons |
Methods for Unfairness |
Regularization |
|
|
2019 |
KDD |
|
Controlling popularity bias in learning-to-rank recommendation |
Methods for Unfairness |
Regularization |
|
|
2017 |
RecSys |
|
Fairness-aware tensor-based recommendation |
Methods for Unfairness |
Regularization |
|
FATR |
2018 |
CIKM |
Pyhton |
Controlling fairness and bias in dynamic learning-to-rank |
Methods for Unfairness |
Regularization |
|
FairCo |
2020 |
SIGIR |
Python |
Censoring representations with an adversary |
Methods for Unfairness |
Adversarial Learning |
|
ALFR |
2016 |
ICLR |
|
“Compositional fairness constraints for graph embeddings |
Methods for Unfairness |
Adversarial Learning |
|
|
2019 |
ICML |
Python |
Privacy-aware recommendation with private-attribute protection using adversarial learning |
Methods for Unfairness |
Adversarial Learning |
|
RAP |
2019 |
WSDM |
|
Fairness in decision-making - the causal explanation formula |
Methods for Unfairness |
Causal Modeling |
|
Ctf-DE/Ctf-IE/Ctf-SE |
2018 |
AAAI |
|
Fair inference on outcomes |
Methods for Unfairness |
Causal Modeling |
|
|
2018 |
AAAI |
C++ |
On discrimination discovery and removal in ranked data using causal graph |
Methods for Unfairness |
Causal Modeling |
|
FRank/FDetect |
2018 |
KDD |
|
Counterfactual fairness |
Methods for Unfairness |
Causal Modeling |
|
counterfactual fairness |
2017 |
arXiv |
Python |
Counterfactual fairness: Unidentification bound and algorithm. |
Methods for Unfairness |
Causal Modeling |
|
|
2019 |
IJCAI |
|
A general knowledge distillation framework for counterfactual recommendation via uniform data |
Methods for Mitigating Loop Effect |
Uniform data. |
|
KDCRec |
2020 |
SIGIR |
Python |
Degenerate feedback loops in recommender systems |
Methods for Mitigating Loop Effect |
Uniform data. |
|
Oracle |
2019 |
AIES |
|
Improving ad click prediction by considering non-displayed events |
Methods for Mitigating Loop Effect |
Uniform data. |
|
|
2019 |
CIKM |
Link |
Predicting counterfactuals from large historical data and small randomized trials |
Methods for Mitigating Loop Effect |
Uniform data. |
|
|
2016 |
WWW |
|
Causal embeddings for recommendation |
Methods for Mitigating Loop Effect |
Uniform data. |
|
CausE |
2018 |
RecSys |
Python |
Influence function for unbiased recommendation |
Methods for Mitigating Loop Effect |
Uniform data. |
|
IF4URec |
2020 |
SIGIR |
|
A contextual-bandit approach to personalized news article recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
LinUCB |
2010 |
WWW |
Python |
Interactive social recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
ISR |
2017 |
CIKM |
|
Factorization bandits for interactive recommendation. |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
FactorUCB |
2017 |
AAAI |
|
Interactive collaborative filtering |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
ICF |
2013 |
CIKM |
Python |
Jointly learning to recommend and advertise |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
RAM |
2020 |
KDD |
|
Stabilizing reinforcement learning in dynamic environment with application to online recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
Robust DQN |
2018 |
KDD |
|
Recommendations with negative feedback via pairwise deep reinforcement learning |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
DEERS |
2018 |
KDD |
|
Drn: A deep reinforcement learning framework for news recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
DDQN |
2018 |
WWW |
|
Large-scale interactive recommendation with tree-structured policy gradient |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
TPGR |
2019 |
AAAI |
Python |
Deep reinforcement learning for page-wise recommendations |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
DeepPage |
2018 |
RecSys |
|
Deep reinforcement learning for list-wise recommendations |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
LIRD |
2019 |
KDD |
Python |
A reinforcement learning framework for explainable recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
|
2018 |
ICDM |
|
When people change their mind: Off-policy evaluation in non-stationary recommendation environments |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
|
2019 |
WSDM |
Python |
Top-k off-policy correction for a reinforce recommender system |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
REINFORCE |
2019 |
WSDM |
Python |
Counterfactual evaluation of slate recommendations with sequential reward interactions |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
RIPS |
2020 |
KDD |
Python |
Off-policy evaluation for slate recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
PI |
2017 |
NIPS |
Python |
Joint policy value learning for recommendation |
Methods for Mitigating Loop Effect |
Reinforcement learning. |
|
Dual Bandit |
2020 |
KDD |
Python |
Debiasing the human-recommender system feedback loop in collaborative filtering |
Methods for Mitigating Loop Effect |
Others |
|
propensity MF |
2019 |
WWW |
|
Deconvolving feedbackloops in recommender systems |
Methods for Mitigating Loop Effect |
Others |
|
Deconvolving feedback |
2016 |
NIPS |
|