skip to main content
10.1145/2645710.2645741acmconferencesArticle/Chapter ViewAbstractPublication PagesrecsysConference Proceedingsconference-collections
research-article

Speeding up the Xbox recommender system using a euclidean transformation for inner-product spaces

Published: 06 October 2014 Publication History

Abstract

A prominent approach in collaborative filtering based recommender systems is using dimensionality reduction (matrix factorization) techniques to map users and items into low-dimensional vectors. In such systems, a higher inner product between a user vector and an item vector indicates that the item better suits the user's preference. Traditionally, retrieving the most suitable items is done by scoring and sorting all items. Real world online recommender systems must adhere to strict response-time constraints, so when the number of items is large, scoring all items is intractable.
We propose a novel order preserving transformation, mapping the maximum inner product search problem to Euclidean space nearest neighbor search problem. Utilizing this transformation, we study the efficiency of several (approximate) nearest neighbor data structures. Our final solution is based on a novel use of the PCA-Tree data structure in which results are augmented using paths one hamming distance away from the query (neighborhood boosting). The end result is a system which allows approximate matches (items with relatively high inner product, but not necessarily the highest one). We evaluate our techniques on two large-scale recommendation datasets, Xbox Movies and Yahoo~Music, and show that this technique allows trading off a slight degradation in the recommendation quality for a significant improvement in the retrieval time.

Supplementary Material

JPG File (p257-sidebyside.jpg)
MP4 File (p257-sidebyside.mp4)

References

[1]
Alexandr Andoni and Piotr Indyk. Near-optimal hashing algorithms for approximate nearest neighbor in high dimensions. In FOCS, pages 459--468, 2006.
[2]
Robert M. Bell and Yehuda Koren. Lessons from the netflix prize challenge. SIGKDD Explor. Newsl., 2007.
[3]
Jon Louis Bentley. Multidimensional binary search trees used for associative searching. Commun. ACM, 18(9):509--517, September 1975.
[4]
Andrei Broder. On the resemblance and containment of documents. In Proceedings of the Compression and Complexity of Sequences 1997, pages 21--29, 1997.
[5]
Moses S. Charikar. Similarity estimation techniques from rounding algorithms. In Proceedings of the Thiry-fourth Annual ACM Symposium on Theory of Computing, pages 380--388, 2002.
[6]
Sanjoy Dasgupta and Yoav Freund. Random projection trees and low dimensional manifolds. In Proceedings of the Fortieth Annual ACM Symposium on Theory of Computing, pages 537--546, 2008.
[7]
Gideon Dror, Noam Koenigstein, and Yehuda Koren. Yahoo! music recommendations: Modeling music ratings with temporal dynamics and item taxonomy. In Proc. 5th ACM Conference on Recommender Systems, 2011.
[8]
Gideon Dror, Noam Koenigstein, Yehuda Koren, and Markus Weimer. The Yahoo! music dataset and KDD-Cup'11. Journal Of Machine Learning Research, 17:1--12, 2011.
[9]
Piotr Indyk and Rajeev Motwani. Approximate nearest neighbors: Towards removing the curse of dimensionality. In Proceedings of the Thirtieth Annual ACM Symposium on Theory of Computing, pages 604--613, 1998.
[10]
Mohammad Khoshneshin and W. Nick Street. Collaborative filtering via euclidean embedding. In Proceedings of the fourth ACM conference on Recommender systems, 2010.
[11]
Noam Koenigstein and Yehuda Koren. Towards scalable and accurate item-oriented recommendations. In Proc. 7th ACM Conference on Recommender Systems, 2013.
[12]
Noam Koenigstein and Ulrich Paquet. Xbox movies recommendations: Variational Bayes matrix factorization with embedded feature selection. In Proc. 7th ACM Conference on Recommender Systems, 2013.
[13]
Noam Koenigstein, Parikshit Ram, and Yuval Shavitt. Efficient retrieval of recommendations in a matrix factorization framework. In CIKM, 2012.
[14]
Yehuda Koren, Robert M. Bell, and Chris Volinsky. Matrix factorization techniques for recommender systems. IEEE Computer, 2009.
[15]
James McNames. A fast nearest-neighbor algorithm based on a principal axis search tree. IEEE Trans. Pattern Anal. Mach. Intell., 23(9):964--976, September 2001.
[16]
Ulrich Paquet and Noam Koenigstein. One-class collaborative filtering with random graphs. In Proceedings of the 22nd international conference on World Wide Web, WWW '13, pages 999--1008, 2013.
[17]
Franco P. Preparata and Michael I. Shamos. Computational Geometry: An Introduction. Springer, 1985.
[18]
Parikshit Ram and Alexander Gray. Maximum inner-product search using cone trees. In SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 2012.
[19]
Parikshit Ram and Alexander G. Gray. Which space partitioning tree to use for search? In Christopher J. C. Burges, Léon Bottou, Zoubin Ghahramani, and Kilian Q. Weinberger, editors, NIPS, pages 656--664, 2013.
[20]
Parikshit Ram, Dongryeol Lee, and Alexander G. Gray. Nearest-neighbor search on a time budget via max-margin trees. In SDM, pages 1011--1022. SIAM / Omnipress, 2012.
[21]
Robert F. Sproull. Refinements to nearest-neighbor searching in k-dimensional trees. Algorithmica, 6(4):579--589, 1991.

Cited By

View all
  • (2024)Faster maximum inner product search in high dimensionsProceedings of the 41st International Conference on Machine Learning10.5555/3692070.3694045(48344-48361)Online publication date: 21-Jul-2024
  • (2024)GUITAR: Gradient Pruning toward Fast Neural RankingProceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval10.1145/3626772.3657728(163-173)Online publication date: 10-Jul-2024
  • (2024)OneSparse: A Unified System for Multi-index Vector SearchCompanion Proceedings of the ACM Web Conference 202410.1145/3589335.3648338(393-402)Online publication date: 13-May-2024
  • Show More Cited By

Index Terms

  1. Speeding up the Xbox recommender system using a euclidean transformation for inner-product spaces

    Recommendations

    Comments

    Information & Contributors

    Information

    Published In

    cover image ACM Conferences
    RecSys '14: Proceedings of the 8th ACM Conference on Recommender systems
    October 2014
    458 pages
    ISBN:9781450326681
    DOI:10.1145/2645710
    Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

    Sponsors

    In-Cooperation

    Publisher

    Association for Computing Machinery

    New York, NY, United States

    Publication History

    Published: 06 October 2014

    Permissions

    Request permissions for this article.

    Check for updates

    Author Tags

    1. fast retrieval
    2. inner product search
    3. matrix factorization
    4. recommender systems

    Qualifiers

    • Research-article

    Conference

    RecSys'14
    Sponsor:
    RecSys'14: Eighth ACM Conference on Recommender Systems
    October 6 - 10, 2014
    California, Foster City, Silicon Valley, USA

    Acceptance Rates

    RecSys '14 Paper Acceptance Rate 35 of 234 submissions, 15%;
    Overall Acceptance Rate 254 of 1,295 submissions, 20%

    Contributors

    Other Metrics

    Bibliometrics & Citations

    Bibliometrics

    Article Metrics

    • Downloads (Last 12 months)43
    • Downloads (Last 6 weeks)9
    Reflects downloads up to 21 Jan 2025

    Other Metrics

    Citations

    Cited By

    View all
    • (2024)Faster maximum inner product search in high dimensionsProceedings of the 41st International Conference on Machine Learning10.5555/3692070.3694045(48344-48361)Online publication date: 21-Jul-2024
    • (2024)GUITAR: Gradient Pruning toward Fast Neural RankingProceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval10.1145/3626772.3657728(163-173)Online publication date: 10-Jul-2024
    • (2024)OneSparse: A Unified System for Multi-index Vector SearchCompanion Proceedings of the ACM Web Conference 202410.1145/3589335.3648338(393-402)Online publication date: 13-May-2024
    • (2024)Fuzzy Norm-Explicit Product Quantization for Recommender SystemsIEEE Transactions on Fuzzy Systems10.1109/TFUZZ.2024.336572232:5(2987-2998)Online publication date: May-2024
    • (2024)Reconsidering Tree based Methods for k-Maximum Inner-Product Search: The LRUS-CoverTree2024 IEEE 40th International Conference on Data Engineering (ICDE)10.1109/ICDE60146.2024.00355(4671-4684)Online publication date: 13-May-2024
    • (2024)Efficient Approximate Maximum Inner Product Search Over Sparse Vectors2024 IEEE 40th International Conference on Data Engineering (ICDE)10.1109/ICDE60146.2024.00303(3961-3974)Online publication date: 13-May-2024
    • (2024)Multi-stage knowledge distillation for sequential recommendation with interest knowledgeInformation Sciences10.1016/j.ins.2023.119841654(119841)Online publication date: Jan-2024
    • (2023)FARGO: Fast Maximum Inner Product Search via Global Multi-ProbingProceedings of the VLDB Endowment10.14778/3579075.357908416:5(1100-1112)Online publication date: 6-Mar-2023
    • (2023)Reverse Maximum Inner Product Search: Formulation, Algorithms, and AnalysisACM Transactions on the Web10.1145/3587215Online publication date: 16-Mar-2023
    • (2023)OPORP: One Permutation + One Random ProjectionProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining10.1145/3580305.3599457(1303-1315)Online publication date: 6-Aug-2023
    • Show More Cited By

    View Options

    Login options

    View options

    PDF

    View or Download as a PDF file.

    PDF

    eReader

    View online with eReader.

    eReader

    Media

    Figures

    Other

    Tables

    Share

    Share

    Share this Publication link

    Share on social media