Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
skip to main content
research-article

Ranking from Crowdsourced Pairwise Comparisons via Smoothed Riemannian Optimization

Published: 09 February 2020 Publication History
  • Get Citation Alerts
  • Abstract

    Social Internet of Things has recently become a promising paradigm for augmenting the capability of humans and devices connected in the networks to provide services. In social Internet of Things network, crowdsourcing that collects the intelligence of the human crowd has served as a powerful tool for data acquisition and distributed computing. To support critical applications (e.g., a recommendation system and assessing the inequality of urban perception), in this article, we shall focus on the collaborative ranking problems for user preference prediction from crowdsourced pairwise comparisons. Based on the Bradley--Terry--Luce (BTL) model, a maximum likelihood estimation (MLE) is proposed via low-rank approach in order to estimate the underlying weight/score matrix, thereby predicting the ranking list for each user. A novel regularized formulation with the smoothed surrogate of elementwise infinity norm is proposed in order to address the unique challenge of the coupled the non-smooth elementwise infinity norm constraint and non-convex low-rank constraint in the MLE problem. We solve the resulting smoothed rank-constrained optimization problem via developing the Riemannian trust-region algorithm on quotient manifolds of fixed-rank matrices, which enjoys the superlinear convergence rate. The admirable performance and algorithmic advantages of the proposed method over the state-of-the-art algorithms are demonstrated via numerical results. Moreover, the proposed method outperforms state-of-the-art algorithms on large collaborative filtering datasets in both success rate of inferring preference and normalized discounted cumulative gain.

    References

    [1]
    P. A. Absil, R. Mahony, and R. Sepulchre. 2009. Optimization Algorithms on Matrix Manifolds. Princeton University Press.
    [2]
    Alekh Agarwal, Sahand Negahban, and Martin J. Wainwright. 2012. Fast global convergence of gradient methods for high-dimensional statistical recovery. Ann. Stat. 40, 5 (Oct. 2012), 2452--2482.
    [3]
    Mine Alsan, Ranjitha Prasad, and Vincent Y. F. Tan. 2018. Lower bounds on the bayes risk of the Bayesian BTL model with applications to comparison graphs. IEEE J. Sel. Topics Signal Process. 12, 5 (Apr. 2018), 975--988.
    [4]
    Suhrid Balakrishnan and Sumit Chopra. 2012. Collaborative ranking. In Proc. ACM Int. Conference on Web Search and Data Mining (WSDM’12). ACM, 143--152.
    [5]
    Afonso S. Bandeira, Nicolas Boumal, and Vladislav Voroninski. 2016. On the low-rank approach for semidefinite programs arising in synchronization and community detection. In Proc. Conf. Learn. Theory (COLT’16). 23--26.
    [6]
    Amir Beck and Marc Teboulle. 2012. Smoothing and first order methods: A unified framework. SIAM J. Optim. 22, 2 (Jun. 2012), 557--580.
    [7]
    Amir Beck and Marc Teboulle. 2003. Mirror descent and nonlinear projected subgradient methods for convex optimization. Oper. Res. Lett. 31, 3 (May. 2003), 167--175.
    [8]
    Sonia A. Bhaskar and Adel Javanmard. 2015. 1-bit matrix completion under exact low-rank constraint. In In Proc. IEEE Conf. Inform. Science and Systems (CISS’15). IEEE, 1--6.
    [9]
    Srinadh Bhojanapalli, Behnam Neyshabur, and Nati Srebro. 2016. Global optimality of local search for low rank matrix recovery. In Adv. Neural. Inf. Process. Syst. (NIPS’16). 3873--3881.
    [10]
    Léon Bottou. 2010. Large-scale machine learning with stochastic gradient descent. In Proc. Int. Conf. Comput. Stat.Springer, 177--186.
    [11]
    Nicolas Boumal and Pierre-Antoine Absil. 2011. RTRMC: A Riemannian trust-region method for low-rank matrix completion. In Adv. Neural. Inf. Process. Syst. (NIPS’11). 406--414.
    [12]
    Nicolas Boumal, P.-A. Absil, and Coralia Cartis. 2018. Global rates of convergence for nonconvex optimization on manifolds. IMA J Numer. Anal. 39, 1 (Feb 2018), 1--33.
    [13]
    Nicolas Boumal, Bamdev Mishra, P.-A. Absil, and Rodolphe Sepulchre. 2014. Manopt, a Matlab toolbox for optimization on manifolds. J. Mach. Learn. Res. 15 (Jan. 2014), 1455--1459. Retrieved from http://www.manopt.org.
    [14]
    Stephen Boyd and Lieven Vandenberghe. 2004. Convex Optimization. Cambridge University Press.
    [15]
    Ralph Allan Bradley and Milton E. Terry. 1952. Rank analysis of incomplete block designs: I. the method of paired comparisons. Biometrika 39, 3-4 (Dec. 1952), 324--345.
    [16]
    Emmanuel Candes and Justin Romberg. 2005. ℓ1-magic: Recovery of sparse signals via convex programming. California Inst. Technol., Pasadena, CA.
    [17]
    Yudong Chen and Yuejie Chi. 2018. Harnessing structures in big data via guaranteed low-rank matrix estimation: Recent theory and fast algorithms via convex and nonconvex optimization. IEEE Signal Process. Mag. 35, 4 (Jul. 2018), 14--31.
    [18]
    Yuxin Chen and Changho Suh. 2015. Spectral MLE: Top-k rank aggregation from pairwise comparisons. In Proc. Int. Conf. Mach. Learn. (ICML’15). 371--380.
    [19]
    Yudong Chen and Martin J. Wainwright. 2015. Fast low-rank estimation by projected gradient descent: General statistical and algorithmic guarantees. arXiv preprint arXiv:1509.03025 (2015).
    [20]
    Mark A. Davenport, Yaniv Plan, Ewout van den Berg, and Mary Wootters. 2014. 1-bit matrix completion. Inf. Inference 3, 3 (Sep. 2014), 189--223.
    [21]
    Mark A. Davenport and Justin Romberg. 2016. An overview of low-rank matrix recovery from incomplete observations. IEEE J. Sel. Topics Signal Process. 10, 4 (Jan. 2016), 608--622.
    [22]
    Wei Feng, Zheng Yan, Hengrun Zhang, Kai Zeng, Yu Xiao, and Y. Thomas Hou. 2018. A survey on security, privacy, and trust in mobile crowdsourcing. IEEE Internet Things J. 5, 4 (Aug. 2018), 2971--2992.
    [23]
    Reinhard Heckel, Max Simchowitz, Kannan Ramchandran, and Martin J. Wainwright. 2018. Approximate ranking from pairwise comparisons. In International Conference on Artificial Intelligence and Statistics. 1057--1066.
    [24]
    Jia Hu, Hui Lin, Xuancheng Guo, and Ji Yang. 2018. DTCS: An integrated strategy for enhancing data trustworthiness in mobile crowdsourcing. IEEE Internet Things J. 5, 6 (Feb. 2018), 4663--4671.
    [25]
    Luis G. Jaimes, Idalides J. Vergara-Laurens, and Andrew Raij. 2015. A survey of incentive techniques for mobile crowd sensing. IEEE Internet Things J. 2, 5 (Oct. 2015), 370--380.
    [26]
    Prateek Jain, Praneeth Netrapalli, and Sujay Sanghavi. 2013. Low-rank matrix completion using alternating minimization. In ACM Symp. Theory Comput. ACM, 665--674.
    [27]
    Sham Kakade, Shai Shalev-Shwartz, and Ambuj Tewari. 2009. On the duality of strong convexity and strong smoothness: Learning applications and matrix regularization. Unpublished Manuscript (2009).
    [28]
    Joonseok Lee, Samy Bengio, Seungyeon Kim, Guy Lebanon, and Yoram Singer. 2014. Local collaborative ranking. In Proc. Int. Conf. World Wide Web. ACM, 85--96.
    [29]
    Jason D. Lee, Max Simchowitz, Michael I. Jordan, and Benjamin Recht. 2016. Gradient descent only converges to minimizers. In 29th Annu. Proc. Conf. Learn. Theory (COLT’16). 1246--1257.
    [30]
    Yu Lu and Sahand N. Negahban. 2015. Individualized rank aggregation using nuclear norm regularization. In Proc. 53rd Annu. Allerton Conf. Commun., Control, Comput. IEEE, 1473--1479.
    [31]
    Gilles Meyer, Silvere Bonnabel, and Rodolphe Sepulchre. 2011. Linear regression under fixed-rank constraints: A Riemannian approach. In Proc. Int. Conf. Mach. Learn. (ICML’11).
    [32]
    Bamdev Mishra, Gilles Meyer, Silv Bonnabel, and Rodolphe Sepulchre. 2014. Fixed-rank matrix factorizations and Riemannian low-rank optimization. Comput. Statist. 29, 3--4 (Jun. 2014), 591--621.
    [33]
    Dohyung Park, Anastasios Kyrillidis, Constantine Caramanis, and Sujay Sanghavi. 2018. Finding low-rank solutions via non-convex matrix factorization, efficiently and provably. SIAM J. Imaging Sci. 11, 4 (Oct. 2018), 2165--2204.
    [34]
    Dohyung Park, Anastasios Kyrillidis, Constantine Caramanis, and Sujay Sanghavi. 2017. Non-square matrix sensing without spurious local minima via the Burer-Monteiro approach. In Proc. Int. Conf. Artificial Intelligence and Statistics (AISTATS’17).
    [35]
    Dohyung Park, Joe Neeman, Jin Zhang, Sujay Sanghavi, and Inderjit S. Dhillon. 2015. Preference completion: Large-scale collaborative ranking from pairwise comparisons. In Proc. Int. Conf. Mach. Learn. (ICML’15). 1907--1916.
    [36]
    Jasson D. M. Rennie and Nathan Srebro. 2005. Fast maximum margin matrix factorization for collaborative prediction. In Proc. Int. Conf. Mach. Learn. (ICML’05). 713--719.
    [37]
    Philip Salesses, Katja Schechtner, and César A. Hidalgo. 2013. The collaborative image of the city: Mapping the inequality of urban perception. PloS One 8, 7 (Jun. 2013), e68400.
    [38]
    Matthew J. Salganik and Karen E. C. Levy. 2015. Wiki surveys: Open and quantifiable social data collection. PloS One 10, 5 (May. 2015), e0123483.
    [39]
    Omer Berat Sezer, Erdogan Dogdu, and Ahmet Murat Ozbayoglu. 2018. Context-aware computing, learning, and big data in Internet of Things: A survey. IEEE Internet Things J. 5, 1 (Feb. 2018), 1--27.
    [40]
    Nihar B. Shah and Martin J. Wainwright. 2017. Simple, robust and optimal ranking from pairwise comparisons. J. Mach. Learn. Res. 18, 199 (Jan. 2017), 1--199.
    [41]
    Ruoyu Sun and Zhi-Quan Luo. 2016. Guaranteed matrix completion via non-convex factorization. IEEE Trans. Inf. Theory 62, 11 (Nov. 2016), 6535--6579.
    [42]
    Alan M. Thompson, John C. Brown, Jim W. Kay, and D. Michael Titterington. 1991. A study of methods of choosing the smoothing parameter in image restoration by regularization. IEEE Trans. Pattern Anal. Mach. Intell. 13, 4 (Apr. 1991), 326--339.
    [43]
    Maksims Volkovs and Richard S. Zemel. 2012. Collaborative ranking with 17 parameters. In Adv. Neural. Inf. Process. Syst. (NIPS’12). 2294--2302.
    [44]
    Kun Wang, Xin Qi, Lei Shu, Der-jiunn Deng, and Joel J. P. C. Rodrigues. 2016. Toward trustworthy crowdsourcing in the social Internet of Things. IEEE Wireless Commun. 23, 5 (Oct. 2016), 30--36.
    [45]
    Markus Weimer, Alexandros Karatzoglou, Quoc V. Le, and Alex J. Smola. 2008. Cofi rank-maximum margin matrix factorization for collaborative ranking. In Adv. Neural. Inf. Process. Syst. (NIPS’08). 1593--1600.
    [46]
    Jun Xu, Wei Zeng, Yanyan Lan, Jiafeng Guo, and Xueqi Cheng. 2018. Modeling the parameter interactions in ranking SVM with low-rank approximation. IEEE Trans. Knowl. Data Eng. 31, 6 (Jun. 2018), 1181--1193.
    [47]
    Peng Yang, Ning Zhang, Shan Zhang, Kan Yang, Li Yu, and Xuemin Shen. 2017. Identifying the most valuable workers in fog-assisted spatial crowdsourcing. IEEE Internet Things J. 4, 5 (Oct. 2017), 1193--1203.

    Cited By

    View all
    • (2024)CrowdDC: Ranking From Crowdsourced Paired Comparison With Divide-and-ConquerIEEE Transactions on Computational Social Systems10.1109/TCSS.2023.329663211:2(3015-3021)Online publication date: May-2024

    Index Terms

    1. Ranking from Crowdsourced Pairwise Comparisons via Smoothed Riemannian Optimization

      Recommendations

      Comments

      Information & Contributors

      Information

      Published In

      cover image ACM Transactions on Knowledge Discovery from Data
      ACM Transactions on Knowledge Discovery from Data  Volume 14, Issue 2
      April 2020
      322 pages
      ISSN:1556-4681
      EISSN:1556-472X
      DOI:10.1145/3382774
      Issue’s Table of Contents
      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 ACM 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]

      Publisher

      Association for Computing Machinery

      New York, NY, United States

      Publication History

      Published: 09 February 2020
      Accepted: 01 November 2019
      Revised: 01 September 2019
      Received: 01 January 2019
      Published in TKDD Volume 14, Issue 2

      Permissions

      Request permissions for this article.

      Check for updates

      Author Tags

      1. Ranking
      2. crowdsourced data
      3. low-rank optimization
      4. pairwise comparison
      5. smoothed matrix manifold optimization
      6. social Internet of Things

      Qualifiers

      • Research-article
      • Research
      • Refereed

      Funding Sources

      • National Nature Science Foundation of China

      Contributors

      Other Metrics

      Bibliometrics & Citations

      Bibliometrics

      Article Metrics

      • Downloads (Last 12 months)15
      • Downloads (Last 6 weeks)4
      Reflects downloads up to 26 Jul 2024

      Other Metrics

      Citations

      Cited By

      View all
      • (2024)CrowdDC: Ranking From Crowdsourced Paired Comparison With Divide-and-ConquerIEEE Transactions on Computational Social Systems10.1109/TCSS.2023.329663211:2(3015-3021)Online publication date: May-2024

      View Options

      Get Access

      Login options

      Full Access

      View options

      PDF

      View or Download as a PDF file.

      PDF

      eReader

      View online with eReader.

      eReader

      HTML Format

      View this article in HTML Format.

      HTML Format

      Media

      Figures

      Other

      Tables

      Share

      Share

      Share this Publication link

      Share on social media