Answering new urban questions: Using eXplainable AI-driven analysis to identify determinants of Airbnb price in Dublin

This paper presents a novel approach for leveraging an eXplainable artificial intelligence approach to examine the price determinants of Airbnb in Dublin, Ireland, mainly focusing on urban characteristics. We employ an interpretable machine learning (ML) approach on publicly accessible Airbnb data to identify optimal Airbnb rental features, predict price trends, and understand Airbnb’s growth in Dublin. Our analysis also integrates features from three main categories: physical geography (e.g., elevation), social geography (e.g., Electoral Division), and infrastructure & economic geography based on housing/rental market theories (e.g., distance from the city centre). This integration of geographical attributes with ML algorithms provides localised insights into Airbnb pricing dynamics, addressing gaps in previous research that often overlooked the impact of geographical features on price prediction. By applying the TreeSHAP utility (tree-based attribute for SHapley Additive exPlanations), we establish the significant role of these geographical features in predicting Airbnb rental prices, thereby reinforcing theories like Alonso’s Bid-Rent Theory. We further validate the robustness of our models through testing procedures. Our study offers considerable value to urban planners, policymakers, urban researchers, and market stakeholders, deepening the understanding of Airbnb’s impact on urban landscapes and contributing to the broader discourse on the Sharing Economy.