Profiling Reviewer Mobility Behavior Using Yelp Reviews

  • Muhammad Usman Zafar FAST School Of Computing NUCES, Chiniot-Faisalabad Campus Chiniot, Pakistan
  • Muhammad Bilal School of Engineering and Technology Sunway University Subang Jaya, Malaysia
  • Nadia Malik Dept. of Software Engineering Faculty of Computing and IT University of Sargodha Sargodha, Pakistan
Keywords: User Profiling, Online Reviews, Yelp, Clustering

Abstract

Social media data has become a popular source for uncovering hidden patterns in user mobility. However, the volume of online customer reviews has increased enormously which makes it difficult for individuals and businesses to identify and analyze these patterns. Moreover, very few studies exist that explore the mobility patterns of reviewers. Therefore, this study aims to profile the mobility of reviewers to identify hidden patterns. The data of 1217 reviewers and 139,187 reviews from Yelp is used in this study. Firstly, distance-based mobility profiles which include the average distance and total distance of reviewers are created. The correlation analysis showed that distance-based measures strongly correlate with the number of reviews written by a reviewer, the number of cities, and the number of states visited by the reviewer compared to other features. Clustering is done using k-means and Density-based spatial clustering of applications with noise (DBSCAN) to analyze the different nature of reviewers which showed that reviewers can be grouped into two categories that are “less traveler” and “more traveler”. A comparison of reviewers that have visited the same number of businesses reveals that their mobility patterns can vary significantly. Finally, the classification of reviewers into less and more travelers is done using various machine learning algorithms. Random forest (RF) achieved Area Under the Curve (AUC) of 0.865, which is comparatively better than other algorithms. The feature importance calculated using RF showed that review count, city, and state are more important features compared to others.

 

Published
2024-05-26
How to Cite
Muhammad Usman Zafar, Muhammad Bilal, & Nadia Malik. (2024). Profiling Reviewer Mobility Behavior Using Yelp Reviews. University of Sindh Journal of Information and Communication Technology , 7(2), 57-64. Retrieved from https://sujo.usindh.edu.pk/index.php/USJICT/article/view/6981

Most read articles by the same author(s)

Obs.: This plugin requires at least one statistics/report plugin to be enabled. If your statistics plugins provide more than one metric then please also select a main metric on the admin's site settings page and/or on the journal manager's settings pages.