IP Library › Granted Patent US 12,530,435
Granted Patent B2
US 12,530,435 · App. 17/975,473 · Granted Jan 20, 2026

One-class recommender system with hinge pairwise distance loss and orthogonal representations

Inventor: Ramin Raziperchikolaei (San Mateo, CA)
Assignee: Rakuten Group, Inc.
G06F18/2433G06F18/214G06F18/22
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,530,435
App. No.
17/975,473
Granted
Jan 20, 2026
Kind
B2
Abstract

Known methods for training a recommender system rely on both similar and dissimilar user-item pairs. Using dissimilar pairs introduces several challenges, such as increasing training time or labeling pairs with unknown interactions as dissimilar even though the user may like the item if presented with it. If only similar pairs are used in the known methods, the result is a collapsed solution in which all users and items are mapped to the same representations. The recommender system and methods disclosed herein overcome these challenges by using only similar pairs but adding two terms to the objective function that prevent a collapsed or partially-collapsed solution. Specifically, the objective function includes a pairwise distance loss term that keeps the average pairwise distance between representations greater than a margin, and an orthogonality loss term that reduces correlations between dimensions in the vector space.

Claims (68)

1 . A method, performed by a computer system, for predicting an interaction value for a user and an item using a one-class recommendation model, the method comprising:

training a one-class recommendation model by performing the following:

(a) obtaining a training dataset that includes user and item data for only similar user-item pairs and excludes dissimilar user-item pairs, wherein:

similar user item pairs are user-item pairs that have a known interaction; and

dissimilar user-item pairs are user-item pairs for which there is no known interaction;

(b) applying the model to the similar user-item pair data in the training dataset to obtain a predicted interaction value for each of the similar user-item pairs, wherein obtaining the predicted interaction value includes generating vector representations in a vector space of the user and items in the training dataset;

(c) calculating a loss value for the model using a loss function that comprises the following loss terms:

an attractive loss term that minimizes a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset;

a pairwise distance loss term that keeps the average pairwise distance in the vector space between all vector representations in the training dataset greater than a margin; and

an orthogonality loss term that reduces correlations between the dimensions in the vector space;

(d) adjusting a set of trainable parameters of the model to reduce the loss value; and

(e) repeating steps (b)-(d) for a number of iterations; and

using the trained model to obtain user-item interaction value predictions with respect to user-item pairs for which there is no known interaction.

2 . The method of claim 1 , wherein the attractive loss term is a contrastive loss based on a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset.

3 . The method of claim 1 , wherein the attractive loss term is a mean squared error loss based on a difference between the predicted interaction value and an actual interaction value for each of the similar user-item pairs within the training dataset.

4 . The method of claim 1 , wherein the pairwise distance loss term is a hinge pairwise distance loss.

5 . The method of claim 1 , wherein the orthogonality loss term makes dimensions in the vector space orthogonal.

6 . The method of claim 1 , wherein the trained model is used to predict user-item interactions for user-item pairs on an ecommerce platform for which no interaction value is known and wherein the method further comprises:

recommending one or more test users on the ecommerce platform to a shop on the platform based on predicted interaction values for an item sold by the shop and each of a plurality of test users.

7 . The method of claim 1 , wherein the trained model is used to predict user-item interactions for user-item pairs on an ecommerce platform for which no interaction value is known and wherein the method further comprises:

recommending one or more items to a test user on the ecommerce platform based on predicted interaction values for the test user and each of a plurality of items on the ecommerce platform.

8 . A method, performed by a computer system, for training a one-class recommendation model to predict interaction values for users and items using training data that consists of only user-item pairs with positive interaction values, the method comprising:

(a) obtaining a training dataset that includes user and item data for only similar user-item pairs and excludes dissimilar user-item pairs, wherein:

similar user item pairs are user-item pairs that have a known interaction; and

dissimilar user-item pairs are user-item pairs for which there is no known interaction;

(b) applying the model to the similar user-item pair data in the training dataset to obtain a predicted interaction value for each of the similar user-item pairs, wherein obtaining the predicted interaction value includes generating vector representations in a vector space of the user and items in the training dataset;

(c) calculating a loss value for the model using a loss function that comprises the following loss terms:

an attractive loss term that minimizes a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset;

a pairwise distance loss term that keeps the average pairwise distance in the vector space between all vector representations in the training dataset greater than a margin; and

an orthogonality loss term that reduces correlations between the dimensions in the vector space;

(d) adjusting a set of trainable parameters of the model to reduce the loss value; and

(e) repeating steps (b)-(d) for a number of iterations.

9 . The method of claim 8 , wherein the attractive loss term is a contrastive loss based on a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset.

10 . The method of claim 8 , wherein the attractive loss term is a mean squared error loss based on a difference between the predicted interaction value and an actual interaction value for each of the similar user-item pairs within the training dataset.

11 . The method of claim 8 , wherein the pairwise distance loss term is a hinge pairwise distance loss.

12 . The method of claim 8 , wherein the orthogonality loss term makes all the vector representations in the training dataset orthogonal.

13 . A system for predicting user-item interaction values on an ecommerce platform, the system comprising:

a processor configured to execute instructions programmed using a set of machine codes;

one or more memory units coupled to the processor; and

a one-class machine-learning recommendation model, stored in the one or more memory units of the system, that takes a user input and an item input and outputs a user-interaction score corresponding to a predicted user-interaction value for users and items on an ecommerce platform, wherein the model includes computational instructions implemented in the machine code for generating the output, and wherein the model is trained according to a method that comprises the following:

(a) obtaining a training dataset that includes user and item data for only similar user-item pairs and excludes dissimilar user-item pairs, wherein:

similar user item pairs are user-item pairs that have a known interaction; and

dissimilar user-item pairs are user-item pairs for which there is no known interaction;

(b) applying the model to the similar user-item pair data in the training dataset to obtain a predicted interaction value for each of the similar user-item pairs, wherein obtaining the predicted interaction value includes generating vector representations in a vector space of the user and items in the training dataset;

(c) calculating a loss value for the model using a loss function that comprises the following loss terms:

an attractive loss term that minimizes a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset;

a pairwise distance loss term that keeps the average pairwise distance in the vector space between all vector representations in the training dataset greater than a margin; and

an orthogonality loss term that reduces correlations between the dimensions in the vector space;

(d) adjusting a set of trainable parameters of the model to reduce the loss value; and

repeating steps (b)-(d) for a number of iterations.

14 . The system of claim 13 , wherein the attractive loss term is a contrastive loss based on a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset.

15 . The system of claim 13 , wherein the attractive loss term is a mean squared error loss based on a difference between the predicted interaction value and an actual interaction value for each of the similar user-item pairs within the training dataset.

16 . The system of claim 13 , wherein the pairwise distance loss term is a hinge pairwise distance loss.

17 . The system of claim 13 , wherein the orthogonality loss term makes all the vector representations in the training dataset orthogonal.

18 . A non-transitory computer-readable medium comprising a computer program, that, when executed by a computer system, enables the computer system to perform the following method for predicting user-item interaction values on an ecommerce platform that includes products from different shops with different sales volumes, the method comprising:

applying a one-class recommendation model to user and item data on an ecommerce platform to obtain predicted user-item interaction values, wherein the one-class recommendation model was trained according to the following method:

(a) obtaining a training dataset that includes user and item data for only similar user-item pairs and excludes dissimilar user-item pairs, wherein:

similar user item pairs are user-item pairs that have a known interaction; and

dissimilar user-item pairs are user-item pairs for which there is no known interaction;

(b) applying the model to the similar user-item pair data in the training dataset to obtain a predicted interaction value for each of the similar user-item pairs, wherein obtaining the predicted interaction value includes generating vector representations in a vector space of the user and items in the training dataset;

(c) calculating a loss value for the model using a loss function that comprises the following loss terms:

an attractive loss term that minimizes a distance in the vector space between the vector representations of the user and the item in each of the similar user-item pairs in the training dataset;

a pairwise distance loss term that keeps the average pairwise distance in the vector space between all vector representations in the training dataset greater than a margin; and

an orthogonality loss term that reduces correlations between the dimensions in the vector space;

(d) adjusting a set of trainable parameters of the model to reduce the loss value; and

repeating steps (b)-(d) for a number of iterations.

19 . The non-transitory computer-readable medium of claim 18 , wherein the pairwise distance loss term is a hinge pairwise distance loss.

20 . The non-transitory computer-readable medium of claim 18 , wherein the orthogonality loss term makes all the vector representations in the training dataset orthogonal.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 27, 2022
From: RAZIPERCHIKOLAEI, RAMIN
To: RAKUTEN GROUP, INC.
Reel/Frame 061567/0924 →
Continuity (2)
Provisional Application 63392826 · Jul 27, 2022
Related Publication 20240037191A1 · Feb 1, 2024
References Cited (187)
US 7987188B2 · Neylon et al. · 2011 [cited by applicant]
US 8386336B1 · Fox et al. · 2013 [cited by applicant]
US 8417713B1 · Blair-Goldensohn et al. · 2013 [cited by applicant]
US 8458054B1 · Thakur · 2013 [cited by applicant]
US 10354184B1 · Vitaladevuni et al. · 2019 [cited by applicant]
US 10614381B2 · Hoffman et al. · 2020 [cited by applicant]
US 10650432B1 · Joseph et al. · 2020 [cited by applicant]
US 10698967B2 · Shen et al. · 2020 [cited by applicant]
US 10769524B1 · Natesh · 2020 [cited by applicant]
US 11004135B1 · Sandler et al. · 2021 [cited by applicant]
US 11361365B2 · Greenwald · 2022 [cited by applicant]
US 11651037B2 · Shi et al. · 2023 [cited by applicant]
US 11669759B2 · Horowitz et al. · 2023 [cited by applicant]
US 12277591B2 · Shi et al. · 2025 [cited by applicant]
US 20010021914A1 · Jacobi et al. · 2001 [cited by applicant]
US 20050162670A1 · Shuler, Jr. · 2005 [cited by applicant]
US 20060155684A1 · Liu et al. · 2006 [cited by applicant]
US 20070046675A1 · Iguchi et al. · 2007 [cited by applicant]
US 20070087756A1 · Hoffberg · 2007 [cited by applicant]
US 20080270363A1 · Hunt et al. · 2008 [cited by applicant]
US 20080294996A1 · Hunt et al. · 2008 [cited by applicant]
US 20080319829A1 · Hunt et al. · 2008 [cited by applicant]
US 20090006156A1 · Hunt et al. · 2009 [cited by applicant]
US 20090018996A1 · Hunt et al. · 2009 [cited by applicant]
US 20090110089A1 · Green · 2009 [cited by applicant]
US 20090125371A1 · Neylon et al. · 2009 [cited by applicant]
US 20090281923A1 · Selinger et al. · 2009 [cited by applicant]
US 20100114933A1 · Murdock et al. · 2010 [cited by applicant]
US 20100268661A1 · Levy et al. · 2010 [cited by applicant]
US 20140104495A1 · Preston et al. · 2014 [cited by applicant]
US 20140195931A1 · Kwon et al. · 2014 [cited by applicant]
US 20140321761A1 · Wang et al. · 2014 [cited by applicant]
US 20140330637A1 · Moran et al. · 2014 [cited by applicant]
US 20140344013A1 · Karty et al. · 2014 [cited by applicant]
US 20140351079A1 · Dong et al. · 2014 [cited by applicant]
US 20150112790A1 · Wolinsky et al. · 2015 [cited by applicant]
US 20150154229A1 · An et al. · 2015 [cited by applicant]
US 20150154508A1 · Chen et al. · 2015 [cited by applicant]
US 20150332374A1 · Fano et al. · 2015 [cited by applicant]
US 20150379732A1 · Sayre, III et al. · 2015 [cited by applicant]
US 20160155173A1 · Isaacson et al. · 2016 [cited by applicant]
US 20160180248A1 · Regan · 2016 [cited by applicant]
US 20160292148A1 · Aley et al. · 2016 [cited by applicant]
US 20170185894A1 · Volkovs et al. · 2017 [cited by applicant]
US 20170193011A1 · Kale et al. · 2017 [cited by applicant]
US 20170193997A1 · Chen et al. · 2017 [cited by applicant]
US 20180040064A1 · Grigg et al. · 2018 [cited by applicant]
US 20180158078A1 · Hsieh et al. · 2018 [cited by applicant]
US 20180204111A1 · Zadeh et al. · 2018 [cited by applicant]
US 20180276710A1 · Tietzen et al. · 2018 [cited by applicant]
US 20180308112A1 · Prentice et al. · 2018 [cited by applicant]
US 20190019016A1 · Ikeda et al. · 2019 [cited by applicant]
US 20190034875A1 · Bryan et al. · 2019 [cited by applicant]
US 20190244270A1 · Kim et al. · 2019 [cited by applicant]
US 20200004835A1 · Ramanath et al. · 2020 [cited by applicant]
US 20200004886A1 · Ramanath et al. · 2020 [cited by applicant]
US 20200005134A1 · Ramanath et al. · 2020 [cited by applicant]
US 20200005149A1 · Ramanath et al. · 2020 [cited by applicant]
US 20200005364A1 · Aznaurashvili et al. · 2020 [cited by applicant]
US 20200175022A1 · Nowozin · 2020 [cited by applicant]
US 20200211065A1 · Govindarajalu et al. · 2020 [cited by applicant]
US 20200380027A1 · Aggarwal et al. · 2020 [cited by applicant]
US 20210004437A1 · Zhang et al. · 2021 [cited by applicant]
US 20210012150A1 · Liu et al. · 2021 [cited by applicant]
US 20210073612A1 · Vahdat et al. · 2021 [cited by applicant]
US 20210081462A1 · Lu et al. · 2021 [cited by applicant]
US 20210097400A1 · Lee · 2021 [cited by applicant]
US 20210110306A1 · Krishnan et al. · 2021 [cited by applicant]
US 20210117839A1 · Kulkarni et al. · 2021 [cited by applicant]
US 20210133846A1 · Xu et al. · 2021 [cited by applicant]
US 20210150337A1 · Raziperchikolaei · 2021 [cited by applicant]
US 20210191990A1 · Shi et al. · 2021 [cited by applicant]
US 20210350393A1 · Dagley et al. · 2021 [cited by applicant]
US 20210382935A1 · Huang et al. · 2021 [cited by applicant]
US 20210383254A1 · Renders et al. · 2021 [cited by applicant]
US 20210397892A1 · Huang et al. · 2021 [cited by applicant]
US 20220114643A1 · Raziperchikolaei · 2022 [cited by applicant]
US 20220155940A1 · Olbrich et al. · 2022 [cited by applicant]
US 20220188371A1 · Kaza et al. · 2022 [cited by applicant]
US 20220207073A1 · Sohail et al. · 2022 [cited by applicant]
US 20220277741A1 · Chaudhary et al. · 2022 [cited by applicant]
US 20220300804A1 · Guan et al. · 2022 [cited by applicant]
US 20220414531A1 · Ong et al. · 2022 [cited by applicant]
US 20230033492A1 · Shi et al. · 2023 [cited by applicant]
US 20230036394A1 · Shi et al. · 2023 [cited by applicant]
US 20230036964A1 · Shi et al. · 2023 [cited by applicant]
US 20230045107A1 · Shi et al. · 2023 [cited by applicant]
US 20230055699A1 · Raziperchikolaei · 2023 [cited by applicant]
US 20230095187A1 · Zanoon et al. · 2023 [cited by applicant]
US 20230095226A1 · Xie et al. · 2023 [cited by applicant]
US 20230281448A1 · Ma · 2023 [cited by examiner]
CN 110019652 · 2019 [cited by applicant]
CN 110309331 · 2019 [cited by applicant]
Agarwal, Pankaj et al., “Personalizing Similar Product Recommendations in Fashion E-commerce”, Jun. 29, 2018, 5 pages. [cited by applicant]
Bhaskar, Karthik Raja Kalaiselvi et al., “Implicit Feedback Deep Collaborative Filtering Product Recommendation System”, Sep. 8, 2020, 10 pages. [cited by applicant]
Extended European Search Report dated Jan. 17, 2024, European Patent Application No. 23186376.2, 12 pages. [cited by applicant]
Liu, Yudan et al. “Real-time Attention Based Look-alike Model for Recommender System”, Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019. [cited by applicant]
Long et al., “Composite Correlation Quantization for Efficient Multimodal Retrieval”, SIGIR '16, Jul. 17-21, 2016, pp. 1-11. [cited by applicant]
Luo, Mi et al., “Metaselector: Meta-Learning for Recommendation with User-Level Adaptive Model Selection”, Proceedings of the Web Conference, 2020, pp. 2507-2513. [cited by applicant]
Ma, Yifei et al., “Temporal-Contextual Recommendation in Real-Time”, KDD '20, Aug. 23-27, 2020, pp. 2291-2299. [cited by applicant]
Meshcheryakov, N. “Machine Learning and Algorithmic Bias: a Basic Qualitative Exploration of AI, Machine Learning, Bias and Regulation”, Mar. 3, 2021, 140 pages. [cited by applicant]
Mooney, Raymond J., et al. “Content-Based Book Recommending Using Learning for Text Categorization”, Proceedings of the Fifth ACM conference on Digital Libraries, 2000. [cited by applicant]
Nahta, Ravi et al., “Embedding metadata using deep collaborative filtering to address the cold start problem for the rating prediction task”, Multimedia Tools and Applications, vol. 80, No. 12, Feb. 18, 2021, pp. 18553-… [cited by applicant]
Pan, Feiyang et al., “Warm up Cold-start Advertisements: Improving CTR Predictions via Learning to Learn ID Embeddings”, Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Informat… [cited by applicant]
Raziperchikolaei, Ramin, et al. “Neural Representations in Hybrid Recommender Systems: Prediction versus Regularization”, Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Informa… [cited by applicant]
Raziperchikolaei, Ramin et al., “Shared Neural Item Representations for Completely Cold Start Problem”, Fifteenth ACM Conference on Recommender Systems, 2021, pp. 422-431. [cited by applicant]
Ricci et al., “Recommender Systems Handbook”, 2011, 845 pages. [cited by applicant]
Salakhutdinov, Russ, et al. “Probabilistic Matrix Factorization”, Advances in Neural Information Processing Systems, 2007, pp. 1-8. [cited by applicant]
Sedhain et al, “AutoRec: Autoencoders Meet Collaborative Filtering”, WWW 2015 Companion, May 18-22, 2015, pp. 1-2. [cited by applicant]
Shi, Shaoyun et al. “Attention-based Adaptive Model to Unify Warm and Cold Starts Recommendation”, Proceedings of the 27th ACM International Conference on Information and Knowledge Management, 2018, pp. 127-136. [cited by applicant]
Slack, Dylan et al., “Fairness Warnings and Fair-MAML: Learning Fairly with Minimal Data”, Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, 2020, pp. 200-209. [cited by applicant]
Strub et al., “Hybrid Recommender System based on Autoencoders”, Workshop on Deep Learning for Recommender Systems, Sep. 2016, pp. 1-7. [cited by applicant]
Su et al., “Deep Joint-Semantics Reconstructing Hashing for Large-Scale Unsupervised Cross-Modal Retrieval”, 2019, pp. 3027-3035. [cited by applicant]
Takács, Gábor, et al. “Matrix Factorization and Neighbor Based Algorithms for the Netflix Prize Problem”, Proceedings of the 2008 ACM Conference on Recommender Systems, 2008, pp. 267-274. [cited by applicant]
Van den Oord, Aaron et al. “Deep content-based music recommendation”, Advances in Neural Information Processing Systems 26 (2013), pp. 1-9. [cited by applicant]
Vartak, Manasi et al. “A Meta-Learning Perspective on Cold-Start Recommendations for Items”, Advances in Neural Information Processing Systems, 2017. [cited by applicant]
Vilalta, Ricardo et al., “A Perspective View and Survey of Meta-Learning”, Artificial Intelligence Review, Sep. 2001, pp. 77-95. [cited by applicant]
Vincent, Pascal et al. “Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion”, Journal of Machine Learning Research, 2010, pp. 3371-3408. [cited by applicant]
Volkovs, Maksims et al. “Dropoutnet: Addressing Cold Start in Recommender Systems”, Advances in Neural Information Processing Systems, 2017, pp. 1-10. [cited by applicant]
Wan et al., “Discriminative Latent Semantic Regression for Cross-Modal Hashing of Multimedia Retrieval”, 2018 IEEE Fourth International Conference on Multimedia Big Data (BigMM), Oct. 21, 2018, pp. 1-7. [cited by applicant]
Wang et al., “Collaborative Deep Learning for Recommender Systems”, KDD '15, Aug. 10-13, 2015, pp. 1235-1244. [cited by applicant]
Wang, Chong et al. “Collaborative Topic Modeling for Recommending Scientific Articles”, Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2011, 9 pages. [cited by applicant]
Wang et al., “Effective Multi-Modal Retrieval based on Stacked Auto-Encoders”, Proceedings of the VLDB Endowment, Sep. 1-5, 2014, pp. 649-660. [cited by applicant]
Wang, Huiwei et al., “ML2E: Meta-Learning Embedding Ensemble for Cold-Start Recommendation”, IEEE Access, Sep. 2020, pp. 165757-165768. [cited by applicant]
Wu et al., “Quantized Correlation Hashing for Fast Cross-Modal Search”, Proceedings of the Twenty-Fourth International Joint Conference on Artificial Intelligence, 2015, pp. 3946-3952. [cited by applicant]
Wu et al., “Unsupervised Deep Hashing via Binary Latent Factor Models for Large-scale Cross-modal Retrieval”, Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, 2018, p. 2854-28… [cited by applicant]
Xue, Hong-Jian, et al. “Deep Matrix Factorization Models for Recommender Systems”, Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence, vol. 17, 2017, pp. 3203-3209. [cited by applicant]
Yang et al., “Shared Predictive Cross-Modal Deep Quantization”, IEEE Transactions on Neural Networks and Learning Systems, vol. 29, No. 11, Nov. 2018, pp. 5292-5303. [cited by applicant]
Yin, Wenpeng “Meta-learning for Few-shot Natural Language Processing: a Survey”, Jul. 2020, 7 pages. [cited by applicant]
Yu, Runsheng et al., “Personalized Adaptive Meta Learning for Cold-Start User Preference Prediction”, 35th AAAI Conference on Artificial Intelligence, Feb. 2021, pp. 10772-10780. [cited by applicant]
Yuan, Bowen et al. “Improving Ad Click Prediction by Considering Non-displayed Events”, Proceedings of the 28th ACM International Conference on Information and Knowledge Management, 2019. [cited by applicant]
Zhang, Shuai et al., “Autosvd++: an Efficient Hybrid Collaborative Filtering Model via Contractive Auto-encoders”, Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Re… [cited by applicant]
Zhang et al., “Collaborative Quantization for Cross-Modal Similarity Search”, 2019, pp. 1-10. [cited by applicant]
Zhang et al., “Large-Scale Multimodal Hashing with Semantic Correlation Maximization”, Association for the Advancement of Artificial Intelligence, 2014, pp. 1-7. [cited by applicant]
Zhang, Yin et al. “A Model of Two Tales: Dual Transfer Learning Framework for Improved Long-tail Item Recommendation”, Proceedings of the Web Conference 2021, pp. 2220-2231. [cited by applicant]
Zhang, Yang et al., “How to Retrain Recommender System? A Sequential Meta-Learning Method”, Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, 2020, pp. 1479… [cited by applicant]
Zhang, Yongfeng, et al. “Joint Representation Learning for Top-N Recommendation with Heterogeneous Information Sources”, Proceedings of the 2017 ACM Conference on Information and Knowledge Management, 2017, pp. 1-10. [cited by applicant]
Zhao, Tong “Learning to Search and Recommend From Users Implicit Feedback”, Aug. 2018, 209 pages. [cited by applicant]
Zhu, Yongchun et al. “Learning to Warm up Cold Item Embeddings for Cold-Start Recommendation with Meta Scaling and Shifting Networks”, Proceedings of the 44th International ACM SIGIR Conference on Research and Developme… [cited by applicant]
Antoniou, +A53:A241 Antreas et al., “How to Train Your MAML”, ICLR 2019. [cited by applicant]
Bansal, Trapit et al., “Learning to Few-Shot Learn Across Diverse Natural Language Classification Tasks”, Proceedings of the 28th International Conference on Computational Linguistics, Dec. 2020, pp. 5108-5123. [cited by applicant]
Barkan, Oren et al. “CB2CF: a Neural Multiview Content-to-Collaborative Filtering Model for Completely Cold Item Recommendations”, Proceedings of the 13th ACM Conference on Recommender Systems, 2019, pp. 1-9. [cited by applicant]
Bianchi, Federico et al., “Fantastic Embeddings and How to Align Them: Zero-Shot Inference in a Multi-Shop Scenario”, SIGIR eCOM '20, Jul. 30, 2020, pp. 1-11. [cited by applicant]
Blei, David M. et al. “Latent Dirichlet Allocation”, Journal of Machine Learning Research, 2003, pp. 993-1022. [cited by applicant]
Bohanec, Marko et al. “Explaining Machine Learning Models in Sales Predictions”, Expert Systems with Applications 71, 2017, pp. 416-428. [cited by applicant]
Bronstein et al., “Data Fusion through Cross-modality Metric Learning using Similarity-Sensitive Hashing”, 2010, pp. 1-8. [cited by applicant]
Cai, Qi et al., “Memory Matching Networks for One-Shot Image Recognition”, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 4080-4088. [cited by applicant]
Cao et al., “Collective Deep Quantization for Efficient Cross-Modal Retrieval”, Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, 2017, pp. 3974-3980. [cited by applicant]
Cao et al., “Deep Visual-Semantic Hashing for Cross-Modal Retrieval”, KDD '16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Aug. 2016, pp. 1445-1454. [cited by applicant]
Cao et al., “Correlation Hashing Network for Efficient Cross-Modal Retrieval”, 2016, pp. 1-12. [cited by applicant]
Chen, Jingyuan, et al. “Attentive Collaborative Filtering: Multimedia Recommendation with Item-and Component-Level Attention”, Proceedings of the 40th International ACM SIGIR Conference on Research and Development in In… [cited by applicant]
Chen, Minmin et al. “Marginalized Denoising Autoencoders for Domain Adaptation”, Proceedings of the 29th International Conference on Machine Learning, 2012. [cited by applicant]
Chen, Zhihong et al. “ESAM: Discriminative Domain Adaptation with Non-Displayed Items to Improve Long-Tail Performance”, Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Informat… [cited by applicant]
Cheng et al., “Wide & Deep Learning for Recommender Systems”, DLRS '16, Sep. 15, 2016, pp. 1-4. [cited by applicant]
Chopra, Sumit et al., “Learning a Similarity Metric Discriminatively, with Application to Face Verification”, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), vol. 1. IEEE, 200… [cited by applicant]
Deng et al., “DeepCF: a Unified Framework of Representation Learning and Matching Function Learning in Recommender System”, 2019, pp. 1-9. [cited by applicant]
Ding et al., “Collective Matrix Factorization Hashing for Multimodal Data”, 2014, pp. 4321-4328. [cited by applicant]
Dong et al., “A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems”, Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, 2017, pp. 1309-1315. [cited by applicant]
Dong, Manqing et al, “MAMO: Memory-Augmented Meta-Optimization for Cold-Start Recommendation”, Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2020. [cited by applicant]
Du, Zhengxiao et al. “Sequential Scenario-Specific Meta Learner for Online Recommendation”, Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019, pp. 2895-2904. [cited by applicant]
Finn, Chelsea et al. “Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks”, Proceedings of the 34th International Conference on Machine Learning, vol. 70, 2017, pp. 1126-1135. [cited by applicant]
Gao, Chen et al. “Cross-domain Recommendation Without Sharing User-relevant Data”, The World Wide Web Conference, 2019, pp. 491-502. [cited by applicant]
Gharibshah, Zhabiz et al., “User Response Prediction in Online Advertising”, ACM Comput. Surv., vol. 37, No. 4, Article 111, Aug. 2021, pp. 1-49. [cited by applicant]
Gharibshah, Zhabiz et al., “Deep Learning for User Interest and Response Prediction in Online Display Advertising”, Data Science and Engineering 5.1, 2020, pp. 12-26. [cited by applicant]
Gong et al., “Learning Binary Codes for High-Dimensional Data Using Bilinear Projections”, 2013, pp. 484-491. [cited by applicant]
Gopalan, Prem et al., “Scalable Recommendation with Hierarchical Poisson Factorization”, UAI, 2015. [cited by applicant]
Guo et al., “DeepFM: a Factorization-Machine based Neural Network for CTR Prediction”, Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence, 2017, pp. 1725-1731. [cited by applicant]
He, et al., “Neural Collaborative Filtering” Proceedings of the 26th International Conference on World Wide Web (WWW'17) [online], Apr. 3, 2017, pp. 173-182. [cited by applicant]
He et al., “Neural Factorization Machines for Sparse Predictive Analytics”, SIGIR '17, Aug. 7-11, 2017, pp. 355-364. [cited by applicant]
He et al., “Outer Product-based Neural Collaborative Filtering”, Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence 2018, pp. 2227-2233. [cited by applicant]
He, Xiangnan, et al. “NAIS: Neural Attentive Item Similarity Model for Recommendation”, IEEE Transactions on Knowledge and Data Engineering, 2018, 13 pages. [cited by applicant]
Hooda, Rahul et al., “Social Commerce Hybrid Product Recommender”, International Journal of Computer Applications, vol. 100, No. 12, Aug. 2014, pp. 43-49. [cited by applicant]
Jiang et al., “Deep Cross-Modal Hashing”, 2017, pp. 3232-3240. [cited by applicant]
Kanagala, Mukhul “Product Recommendation System Using Machine Learning Techniques”, California State University San Marcos, Dec. 10, 2020, pp. 1-32. [cited by applicant]
Koren, “Factorization Meets the Neighborhood: a Multifaceted Collaborative Filtering Model”, KDD 2008, Aug. 24-27, 2008, pp. 426-434. [cited by applicant]
Koren, Yehuda, et al. “Matrix Factorization Techniques for Recommender Systems”, Computer, Published by IEEE Computer Society, 2009, pp. 42-49. [cited by applicant]
Krishnan, Adit et al., “An Adversarial Approach to Improve Long-Tail Performance in Neural Collaborative Filtering”, Proceedings of the 27th ACM International Conference on Information and Knowledge Management, 2018, pp… [cited by applicant]
Kumar et al., “Learning Hash Functions for Cross-View Similarity Search”, 2011, pp. 1-6. [cited by applicant]
Lee, Hoyeop et al. “MeLU: Meta-Learned User Preference Estimator for Cold-Start Recommendation”, Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019, pp. 1073-1082. [cited by applicant]
Li et al., “Deep Binary Reconstruction for Cross-modal Hashing”, MM '17, Oct. 23-27, 2017, pp. 1-8. [cited by applicant]
Li et al., “Deep Collaborative Filtering via Marginalized Denoising Auto-encoder”, CIKM '15, Oct. 19-23, 2015, pp. 811-820. [cited by applicant]
Li et al., “Coupled Cycle-GAN: Unsupervised Hashing Network for Cross-Modal Retrieval”, Thirty-Third AAAI Conference on Artificial Intelligence, 2019, pp. 176-183. [cited by applicant]
Li et al., “Deep Heterogeneous Autoencoders for Collaborative Filtering”, 2018, pp. 1-6. [cited by applicant]
Li et al., “Self-Supervised Adversarial Hashing Networks for Cross-Modal Retrieval”, 2018, pp. 4242-4251. [cited by applicant]
Lian et al., “xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems”, KDD 2018, Aug. 19-23, 2018, pp. 1-10. [cited by applicant]
Linden, Greg et al., “Amazon.com Recommendations: Item-to-Item Collaborative Filtering”, IEEE Internet Computing, 2003, pp. 76-80. [cited by applicant]
Liu et al., “Recommender Systems with Heterogeneous Side Information”, WWW '19, May 13-17, 2019, pp. 1-7. [cited by applicant]