Proceedings of the 6th Computer Science Research Days, JRI 2023, 18-20 December 2023, Ouagadougou, Burkina Faso

Research Article

Hotels choice prediction in security crisis times in Burkina Faso using exogenous factors, machine learning and Multi-Criteria Optimisation

Download77 downloads
  • @INPROCEEDINGS{10.4108/eai.18-12-2023.2348129,
        author={Mamadou Diarra and Issiaka Sanou and Abdoulaye Sere},
        title={Hotels choice prediction in security crisis times in Burkina Faso using exogenous factors, machine learning and Multi-Criteria Optimisation},
        proceedings={Proceedings of the 6th Computer Science Research Days, JRI 2023, 18-20 December 2023, Ouagadougou, Burkina Faso},
        publisher={EAI},
        proceedings_a={JRI},
        year={2024},
        month={6},
        keywords={hotel market external environmental factors prediction machine learning multi-criteria optimisation},
        doi={10.4108/eai.18-12-2023.2348129}
    }
    
  • Mamadou Diarra
    Issiaka Sanou
    Abdoulaye Sere
    Year: 2024
    Hotels choice prediction in security crisis times in Burkina Faso using exogenous factors, machine learning and Multi-Criteria Optimisation
    JRI
    EAI
    DOI: 10.4108/eai.18-12-2023.2348129
Mamadou Diarra1,*, Issiaka Sanou2, Abdoulaye Sere2
  • 1: Ecole Polytechnique de Ouagadougou
  • 2: Université Nazi Boni
*Contact email: diarra.md21@gmail.com

Abstract

Customer satisfaction and retention are two priorities for the hotel industry. To achieve these objectives, hotels need to offer quality services. However, due to the security situation in Burkina Faso, external factors have become crucial in selecting accommodation sites. So, predictive analysis of data related to these exogenous factors is therefore becoming imperative for decision support. Our decision for this predictive analysis was to compare some classification algorithms and multicriteria optimization. The simulation of different classification algorithms and multi-criteria optimization has shown that exogenous factors have a significant impact on customer choice. The results achieved an average accuracy of 80%.