Characterizing and Predicting Yelp Users’ Behavior

  • Parvathy Jayaprakasan
  • R. N. UmaEmail author
  • A. Sankarasubramanian
Part of the Studies in Big Data book series (SBD, volume 27)


A business’ revenue is significantly dependent on Yelp user ratings (Luca, Reviews, reputation, and revenue: the case of Harvard Business School Working Paper, No. 12-016, 2016; Anderson and Magruder J Econ J. 122(563):957–989, 2012). Knowing the characteristics of Yelp users will influence their business practices that would eventually help improve their Yelp ratings and consequently their revenue. We categorize Yelp users based on the average number of stars given by each user for their reviews. We determine the common characteristics and differences of users between these user groups; and determine whether these characteristics change by business category. We conclude that users whose average rating falls between 3.7 and 4.0 are the most influential and socially connected and that the type of business does not affect the characteristics of the users. Additionally, we design a two-stage predictive model to predict the average star rating of users given their features or attributes and compare its performance to standard models such as random forest and generalized additive model.


Yelp Users Average Star Rating Generalized Additive Models (GAM) Business Category Random Forest Technique 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Anderson M, Magruder J. Learning from the crowd: regression discontinuity estimates of the effects of an online review database. Econ J. 2012; 122(563):957–89.
  2. 2.
    Bhoompally R. Analysis of business ranking for a connected group of Yelp users by aggregating preference Pairs. M.S. Thesis, University of Cincinnati; 2015.Google Scholar
  3. 3.
    Breiman L. Random forests. Mach Learn. 2001;45(1):5–32.CrossRefGoogle Scholar
  4. 4.
    Cleveland WS, Devlin SJ. Locally-weighted regression: an approach to regression analysis by local fitting. J Am Stat Assoc. 1988; 83(403):596–610.CrossRefGoogle Scholar
  5. 5.
    Feng H, Qian X, Recommendation via user’s personality and social contextual. In: ACM international conference on information and knowledge management (CIKM); 2013 Oct-Nov. p. 1521–24.Google Scholar
  6. 6.
    Hastie T, Tibshirani R, Generalized additive models. Stat Sci. 1986;1(3):297–318.CrossRefGoogle Scholar
  7. 7.
    Ho TK. Random decision forests. In: Proceedings of the 3rd international conference on document analysis and recognition, Montreal, QC; 1995 August, p. 278–82.Google Scholar
  8. 8.
    Jindal T. Finding local experts from Yelp dataset. M.S. Thesis, University of Illinois at Urbana-Champaign; 2015.Google Scholar
  9. 9.
    Koenker R, Hallock KF. Quantile regression. J Econ Perspect. 2001; 15(4):143–56.CrossRefGoogle Scholar
  10. 10.
    Kotzias D, Denil M, de Freitas N, Smyth P. From group to individual labels using deep features. In: ACM SIGKDD conference on knowledge discovery and data mining (KDD); 2015 Aug.Google Scholar
  11. 11.
    Luca M. Reviews, reputation, and revenue: the case of Harvard Business School Working Paper, No. 12-016; 2016 March (Revise and resubmit at the American Economic Journal - Applied Economics).
  12. 12.
    Pranata I, Susilo W. Are the most popular users always trustworthy? the case of Yelp. Electron Commer Res Appl. 2016; 20:30–41.CrossRefGoogle Scholar
  13. 13.
  14. 14.
    Zhao G, Qian X, Xie X. User-service rating prediction by exploring social users’ rating behaviors. IEEE Trans Multimedia. 2016;18(3):496–506.CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2018

Authors and Affiliations

  • Parvathy Jayaprakasan
    • 1
  • R. N. Uma
    • 1
    Email author
  • A. Sankarasubramanian
    • 2
  1. 1.Department of Mathematics and PhysicsNorth Carolina Central UniversityDurhamUSA
  2. 2.Department of Civil, Construction and Environmental EngineeringNorth Carolina State UniversityRaleighUSA

Personalised recommendations