29,966 Hits in 5.5 sec

Improving the Quality of Recommendations for Users and Items in the Tail of Distribution

Liang Hu, Longbing Cao, Jian Cao, Zhiping Gu, Guandong Xu, Jie Wang
2017 ACM Transactions on Information Systems  
To improve the quality of recommendations for items and users in the tail of distribution, we propose a coupled regularization approach that consists of two latent factor models: C-HMF, for enhancing credibility  ...  Yet, although the number of long-tail items and users is much larger than that of short-head items and users, in reality, the amount of data associated with long-tail items and users is much less.  ...  -Update parameters {u S i , λ S i } of the distribution Q(U S i ) in parallel, for each i: 45 ) 45 Improving the Quality of Recommendations for Users and Items in the Tail of Distribution Run k-step  ... 
doi:10.1145/3052769 fatcat:5ajw3tf6dbfeznv2udy24cboli

A Model of Two Tales: Dual Transfer Learning Framework for Improved Long-tail Item Recommendation [article]

Yin Zhang, Derek Zhiyuan Cheng, Tiansheng Yao, Xinyang Yi, Lichan Hong, Ed H. Chi
2021 arXiv   pre-print
To improve tail-item recommendation, we conduct research to transfer knowledge from head items to tail items, leveraging the rich user feedback in head items and the semantic connections between head and  ...  The two types of transfers are incorporated to ensure the learned knowledge from head items can be well applied for tail item representation learning in the long-tail distribution settings.  ...  ACKNOWLEDGMENTS The authors would like to thank Maciej Kula, Jiaxi Tang, Wang-Cheng Kang for their help in this work.  ... 
arXiv:2010.15982v2 fatcat:ef436novm5clxmbem35rkkf4wi

Co-training Disentangled Domain Adaptation Network for Leveraging Popularity Bias in Recommenders

Zhihong Chen, Jiawei Wu, Chenliang Li, Jingxu Chen, Rong Xiao, Binqiang Zhao
2022 Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval  
For long-tail distribution shift, we introduce additional unexposed items (most are long-tail items) to align the distribution of hot and long-tail item property representations.  ...  From the long-tail distribution shift perspective, the sparse interactions of long-tail items lead to insufficient learning of them.  ...  For the improvement of the performance in the long-tail set, we owe it to the 𝐿 𝑙𝑜𝑛𝑔−𝑡𝑎𝑖𝑙 , which reduces the disparity between hot and long-tail representation distributions, and makes sufficient  ... 
doi:10.1145/3477495.3531952 fatcat:q35tkhizybdpbmttlfefxwodle

Connecting User and Item Perspectives in Popularity Debiasing for Collaborative Recommendation [article]

Ludovico Boratto, Gianni Fenu, Mirko Marras
2020 arXiv   pre-print
This can hamper user interest and several qualities of the recommended lists (e.g., novelty, coverage, diversity), impacting on the future success of the platform.  ...  To mitigate the influence of popularity, we propose an in-processing approach aimed at minimizing the correlation between user-item relevance and item popularity, leading to a more equal treatment of items  ...  Acknowledgments This work has been partially supported by the Agència per a la Competivitat de l'Empresa, ACCIÓ, under "Fair and Explainable Artificial Intelligence (FX-AI)" Project.  ... 
arXiv:2006.04275v1 fatcat:mhkobdmtkvbehmyrui7u4gmjpi

A Survey of Long-Tail Item Recommendation Methods

Jing Qin, Danfeng Hong
2021 Wireless Communications and Mobile Computing  
The long-tail item recommendation method not only considers the recommendation of short-head items but also considers recommending more long-tail items to users, thus improving the coverage and diversity  ...  of the research on long-tail item recommendation methods (from clustering-based, which began in 2008, to deep learning-based methods, which began in 2020) and the future directions associated with this  ...  Acknowledgments The author would like to thank the authors of all the references.  ... 
doi:10.1155/2021/7536316 fatcat:3in4tt3ntng6lew6gkvysz3rsi

Flatter is better: Percentile Transformations for Recommender Systems [article]

Masoud Mansoury, Robin Burke, Bamshad Mobasher
2019 arXiv   pre-print
This transformation flattens the rating distribution, better compensates for differences in rating distributions, and improves recommendation performance.  ...  It is well known that explicit user ratings in recommender systems are biased towards high ratings, and that users differ significantly in their usage of the rating scale.  ...  In this methodology, for each user in test set, a list of items will be recommended, and then ranking quality will be measured only on long-tail items in the recommended list.  ... 
arXiv:1907.07766v1 fatcat:pmc2txxfa5dddowwhe3o7pplci

FairMatch: A Graph-based Approach for Improving Aggregate Diversity in Recommender Systems [article]

Masoud Mansoury, Himan Abdollahpouri, Mykola Pechenizkiy, Bamshad Mobasher, Robin Burke
2020 arXiv   pre-print
That leads to low coverage of items in recommendation lists across users (i.e. low aggregate diversity) and unfair distribution of recommended items.  ...  The algorithm iteratively finds items that are rarely recommended yet are high-quality and add them to the users' final recommendation lists.  ...  In another work [24] , they proposed the idea of recommending users to items for improving novelty and aggregate diversity.  ... 
arXiv:2005.01148v1 fatcat:3sapsdzrhjcnbkyhhbm3ylbxni

Niche Product Retrieval in Top-N Recommendation

Mi Zhang, Neil Hurley
2010 2010 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology  
A challenge for personalised recommender systems is to target products in the long tail. That is, to recommend products that the end-user likes, but that are not generally popular.  ...  Given a prior probability distribution of relevance based on item popularity, and a user-specific relevance probability, the other strategy uses a number of scores based on distance measures between these  ...  To observe the user preference distribution, we obtain the percentage of items in P u that fall in the head, mid and tail respectively.  ... 
doi:10.1109/wi-iat.2010.79 dblp:conf/webi/ZhangH10 fatcat:6bbjn7qgjvgefnw5lpmf4m3q4u

Analysis of cold-start recommendations in IPTV systems

Paolo Cremonesi, Roberto Turrin
2009 Proceedings of the third ACM conference on Recommender systems - RecSys '09  
profile. new item: when a new item is added to the catalog, it has no ratings. new system: when bootstrapping a new recommender system, the average number of ratings per user and item is low and this  ...  In this paper we evaluate the performance of different collaborative algorithms in cold-start situations, where the initial lack of ratings may affect the quality of the algorithms.  ...  The kNN approach discards the noise of the items poorly correlated to the target item, improving the quality of recommendations.  ... 
doi:10.1145/1639714.1639756 dblp:conf/recsys/CremonesiT09 fatcat:7t5faz5grba7jhe6pne2ujanpm

Serendipitous Personalized Ranking for Top-N Recommendation

Qiuxia Lu, Tianqi Chen, Weinan Zhang, Diyi Yang, Yong Yu
2012 2012 IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology  
However, due to the imbalance in observed data for popular and tail items, existing collaborative filtering methods fail to give satisfactory serendipitous recommendations.  ...  The experimental results demonstrate that our method significantly improves both accuracy and serendipity for top-N recommendation compared to traditional personalized ranking methods in various settings  ...  In fact, serendipitous recommendation has been shown to benefit both e-retailers and users by suggesting both unexpected and useful items, which are always located in the tail of the popularity distribution  ... 
doi:10.1109/wi-iat.2012.135 dblp:conf/webi/LuCZYY12 fatcat:3i2olk2ztnbqdo4zncmhq4nqki

Managing Popularity Bias in Recommender Systems with Personalized Re-ranking [article]

Himan Abdollahpouri, Robin Burke, Bamshad Mobasher
2019 arXiv   pre-print
We also examine both new and existing metrics to measure the coverage of long-tail items in the recommendation.  ...  However, recommending the ignored products in the 'long tail' is critical for businesses as they are less likely to be discovered.  ...  L u is the recommended list of items for user u and |U t | is the number of users in the test set.  ... 
arXiv:1901.07555v4 fatcat:zmlktwsmr5fhrcwgfrwab6sxyi

Improving Aggregate Recommendation Diversity Using Ranking-Based Techniques

G. Adomavicius, YoungOk Kwon
2012 IEEE Transactions on Knowledge and Data Engineering  
In this paper, we introduce and explore a number of item ranking techniques that can generate recommendations that have substantially higher aggregate diversity across all users while maintaining comparable  ...  Recommender systems are becoming increasingly important to individual users and businesses for providing personalized recommendations.  ...  ACKNOWLEDGMENT The research reported in this paper was supported in part by the National Science Foundation grant IIS-0546443.  ... 
doi:10.1109/tkde.2011.15 fatcat:cuslmyhxvjhh3mkzppo6utdewu

CMBPR: Category-Aided Multi-Channel Bayesian Personalized Ranking for Short Video Recommendation

Xichen Wang, Chen Gao, Jingtao Ding, Yong Li, Depeng Jin
2019 IEEE Access  
algorithms and solves the influence of the "Long Tail" effect.  ...  Therefore, short video recommendation is one of the most important research topics in social media.  ...  In order to improve the data quality, we remove the users who watch short videos no more than twice, then we obtain a dataset consisting of about 10k users and 5k short videos.  ... 
doi:10.1109/access.2019.2907494 fatcat:drmd2gsp5jgnphhbqfhquqhr44

GetJar mobile application recommendations with very sparse datasets

Kent Shi, Kamal Ali
2012 Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '12  
For app usage, we observe a distribution that has higher kurtosis (heavier head and longer tail) than that for the aforementioned movie datasets.  ...  The Netflix competition of 2006 [2] has spurred significant activity in the recommendations field, particularly in approaches using latent factor models [3, 5, 8, 12] .  ...  ACKNOWLEDGEMENTS The authors would like to thank Anand Venkataraman for guidance, edits and help with revisions.  ... 
doi:10.1145/2339530.2339563 dblp:conf/kdd/ShiA12 fatcat:ginlocsn7ng2xltnrcoad2uxry

Self-supervised Learning for Large-scale Item Recommendations [article]

Tiansheng Yao, Xinyang Yi, Derek Zhiyuan Cheng, Felix Yu, Ting Chen, Aditya Menon, Lichan Hong, Ed H. Chi, Steve Tjoa, Jieqi Kang, Evan Ettinger
2021 arXiv   pre-print
However, with millions to billions of items in the corpus, users tend to provide feedback for a very small set of them, causing a power-law distribution.  ...  This makes the feedback data for long-tail items extremely sparse.  ...  We observe that the proposed SSL methods improve the performance for both head and tail item recommendations, with larger gains from the tail items.  ... 
arXiv:2007.12865v4 fatcat:euu7phtharckdbwki3cfceqmq4
« Previous Showing results 1 — 15 out of 29,966 results