Title: Location recommendations from user mobile trajectories
Language: English
Authors: Weydemann, Leonard Felix Amedeo 
Qualification level: Diploma
Advisor: Werthner, Hannes 
Assisting Advisor: Sacharidis, Dimitrios  
Issue Date: 2020
Weydemann, L. F. A. (2020). Location recommendations from user mobile trajectories [Diploma Thesis, Technische Universität Wien]. reposiTUm. https://doi.org/10.34726/hss.2020.59222
Number of Pages: 74
Qualification level: Diploma
Location Recommenders (LR) are location based recommender systems which use locations from historic user trajectories to recommend new places such as locations of sights, museums, or other places of interest by learning past preferences. In this work we analyze the viability of different LR models to be used with mobile hotspots which are handed out to tourists in different hotels in the city of Vienna. As these devices can be potentially used by thousands of users at any given time, we explore different methods of data collection, management and visualization. Whenever systems are used in assisting humans with their decision processes, the implications of algorithmically made recommendations need to be analyzed. For LR fairness can be defined for both users and locations, as both can be subject to unfairness. We propose a framework in which multiple fairness aspects can be formulated. We compare the LR in three fairness aspects formulated in the framework and their predictive accuracy, to give a final assessment of which LR is preferable in this scenario.
Keywords: Geografische Empfehlungssysteme; Empfehlungssysteme; Fairness
Location Recommenders; Recommender Systems; Fairness
URI: https://resolver.obvsg.at/urn:nbn:at:at-ubtuw:1-138761
DOI: 10.34726/hss.2020.59222
Library ID: AC15665919
Organisation: E194 - Institut für Information Systems Engineering 
Publication Type: Thesis
Appears in Collections:Thesis

Files in this item:

Google ScholarTM


Items in reposiTUm are protected by copyright, with all rights reserved, unless otherwise indicated.