The transformative power of recommender systems in enhancing citizens’ satisfaction: Evidence from the Moroccan public sector

  • 162 Views
  • 36 Downloads

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License

The study aims to specifically evaluate the potential impact of implementing AI-powered recommender systems on citizen satisfaction within Moroccan public services. As part of its ambitious digital transformation, Morocco is integrating digital technologies into its public sector to enhance service delivery. Recommender systems, by providing personalized, timely, and relevant recommendations, are hypothesized to significantly increase citizens’ satisfaction and transform public service delivery. The study highlights a comprehensive model that captures the complex and interrelated factors influencing recommender system success. This model was tested using Smart PLS (Partial Least Squares) on data collected from a diverse sample of 157 Moroccan citizens. These participants were randomly selected from various demographics and regions to represent the general population’s perspectives on the future implementation of AI-powered recommender systems in public services. The survey tested three hypotheses: the positive relationship between the potential use of recommender systems and anticipated citizen satisfaction (supported; b = 0.694, p = 0.000, t = 21.214), the impact of trust in AI-powered recommender systems on anticipated citizens’ satisfaction (supported; b = 0.543, p = 0.000, t = 14.230) ; and the moderating effect of trust on AI-powered recommender systems showing a positive effect on anticipated satisfaction (supported; b = 0.154, p = 0.000, t = 4.907). These findings suggest that the future integration of AI-powered recommender systems into public services can enhance citizens’ satisfaction, particularly where there is high trust in the technology.

Acknowledgment
This paper is partly supported by Sidi Mohamed ben Abdellah University, Morocco.

view full abstract hide full abstract
    • Figure 1. Research model
    • Figure 2. Results of the conceptual model
    • Table 1. Sample distribution across gender and age
    • Table 2. Age and qualification level distribution
    • Table 3. Hypotheses testing
    • Table 4. Standardized loadings, reliability, and validity
    • Table 5. Fornell-Larcker criterion for discriminant validity
    • Table 6. Standardized root mean squared residual
    • Table 7. HTMT criterion test result
    • Table A1. Use of recommender systems, trust in AI-powered recommender systems and citizens’ satisfactions measures
    • Conceptualization
      Ouissale El Gharbaoui, Hayat El Boukhari, Abdelkader Salmi
    • Data curation
      Ouissale El Gharbaoui, Abdelkader Salmi
    • Formal Analysis
      Ouissale El Gharbaoui, Abdelkader Salmi
    • Investigation
      Ouissale El Gharbaoui, Hayat El Boukhari
    • Methodology
      Ouissale El Gharbaoui, Hayat El Boukhari
    • Project administration
      Ouissale El Gharbaoui, Hayat El Boukhari
    • Software
      Ouissale El Gharbaoui, Abdelkader Salmi
    • Validation
      Ouissale El Gharbaoui, Hayat El Boukhari, Abdelkader Salmi
    • Visualization
      Ouissale El Gharbaoui, Abdelkader Salmi
    • Writing – original draft
      Ouissale El Gharbaoui
    • Writing – review & editing
      Ouissale El Gharbaoui
    • Supervision
      Hayat El Boukhari