A critical study on Yelp dataset for recommender system

This research critically examines the Yelp dataset within the domain of recommender systems, emphasizing the integration of geographical data to enhance recommendation accuracy and relevance. Acknowledging the pivotal role of location in user preferences, this study introduces a novel method for...

Full description

Saved in:
Bibliographic Details
Main Author: Ruan, Donglin
Other Authors: Sun Aixin
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2024
Subjects:
Online Access:https://hdl.handle.net/10356/175307
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:This research critically examines the Yelp dataset within the domain of recommender systems, emphasizing the integration of geographical data to enhance recommendation accuracy and relevance. Acknowledging the pivotal role of location in user preferences, this study introduces a novel method for constructing user-specific bounding boxes that encapsulate the geographical extent of user interactions within the Yelp environment. Through the meticulous construction and adjustment of these bounding boxes based on user review history and ground truth data, we aim to demonstrate the significance of spatial information in refining recommendation systems. Our methodology extends beyond traditional recommendation algorithms by incorporating spatial constraints, thereby offering a more personalized and contextually relevant user experience. Experimental results underscore the efficacy of our approach, revealing a marked improvement in recommendation performance when geographical considerations are factored into the model. This investigation not only sheds light on the untapped potential of geographical data in enhancing recommender systems but also sets the stage for future research endeavors focused on the confluence of spatial information and user preferences. Ultimately, this study contributes to the evolving landscape of recommendation systems, advocating for a more nuanced and comprehensive approach to user personalization.