Understanding the Impact of Multimodal Models on Airbnb Pricing: Evaluating Explainability and Integration

Pandey, Aditya (2025) Understanding the Impact of Multimodal Models on Airbnb Pricing: Evaluating Explainability and Integration. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (4MB) | Preview
Preview	PDF (Configuration Manual) Download (3MB) | Preview

Abstract

The sharing economy relies on dynamic pricing. This approach allows Airbnb hosts to maximize revenue and occupancy rates with regard to various factors in the market. This research introduces a multimodal framework for Airbnb price prediction, integrating tabular property data with guest review text. Based on a listings/reviews data set of 6481/293744 rows, the framework consists of an ensemble of tree-based models (RandomForest, GradientBoosting, and ExtraTrees) as well as DistilBERT embeddings of the reviews and a RandomForest meta-learner that combines these results. Extensive feature engineering with derived metrics such as price-per-person and neighborhood popularity, as well as effective preprocessing (standard scaling, power transformations, and one-hot encoding), gives a representation of all data. SHAP-based explainability highlights key pricing factors, such as location and amenities, offering actionable insights for hosts. The multimodal model achieves an R2 of 0.8641 and MAE of €0.15 (log-price units, €23.42), improving by 0.66% in R2 and 6.25% in MAE over the tabular baseline model (R2: 0.8575, MAE: €0.16 log-price units, €25.87), respectively, through 8-fold cross-validation training. This paper presents a contribution in data analytics because it shows the effectiveness of multimodal combination and explainable AI in pricing analysis.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Gyamfi, Eric UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Hospitality Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning G Geography. Anthropology. Recreation > GF Human ecology. Anthropogeography > Human settlements > Housing > Rental housing
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	02 Jul 2026 14:21
Last Modified:	02 Jul 2026 14:21
URI:	https://norma.ncirl.ie/id/eprint/9444

Actions (login required)

View Item