Deciphering the Airbnb market: DaSCI-UGR researchers use AI to analyse the Canary Islands’ accommodation sector

17 March, 2025

According to the results of the study, accommodations managed by large operators show a lower economic performance compared to individual properties.

Researchers from the DaSCI institute of the University of Granada, together with scientists from the University of Las Palmas de Gran Canaria, have developed a decision support system (DSS) based on social network analysis and data visualisation.

Applied to the holiday accommodation market in the Canary Islands, this model identifies patterns of differentiation and profitability in the P2P sector, providing valuable information for investors and tourism managers, but also for political and social decision-makers.

The study, published in the journal Socio-Economic Planning Sciences, analyses more than 9,000 tourist accommodations in the Canary Islands with real data from Airbnb. The methodology used allows the construction of a visual map that positions each accommodation according to its similarity to others, thus facilitating the identification of key patterns.

Main findings

The researchers identified nine accommodation typologies differentiated by characteristics such as guest capacity, number of properties managed by the host and cancellation policies. Among the most relevant results, they highlight that higher income accommodations tend to be located in peripheral areas of the visual map, indicating that they possess distinctive characteristics. In particular, they have higher accommodation capacity and require longer minimum stays.

However, accommodation managed by large operators shows a lower economic return compared to individual properties. Beyond capacity, the results suggest that strategies such as changing the cancellation policy and obtaining the ‘Superhost’ badge can improve profitability.

‘Differentiation is key in the vacation rental market,’ explains Víctor A. Vargas Pérez, lead author of the study, from the UGR’s Department of Computer Science and Artificial Intelligence. ‘Our system allows us to intuitively visualise which accommodations are most successful and how they are distributed in the market, facilitating strategic decision-making.

Applications for investors and tourism managers

This model not only provides valuable information for investors seeking to maximise their returns, but can also be useful for tourism managers and urban planners. By providing a detailed view of the market, the system helps to identify investment opportunities and design strategies to improve the competitiveness of the sector, as well as housing policies compatible with the local population.

The study emphasises that the classification of accommodation is not linked to a specific location within the archipelago, but that the different typologies are distributed homogeneously across the islands. This finding reinforces the idea that investment strategies should focus on differentiation and management characteristics, rather than relying solely on location.

The research team hopes that the methodology can be applied to other tourism destinations and holiday rental platforms, contributing to the development of more sophisticated analysis tools for the tourism industry.

DaSCI

The Andalusian Inter-University Institute in Data Science and Computational Intelligence, known as DaSCI, is a collaborative entity between the universities of Granada, Jaén and Córdoba. It is dedicated to advanced research and training in the field of artificial intelligence, with a particular focus on data science and computational intelligence.

The institute brings together an outstanding group of researchers working on joint projects, promoting the development and application of innovative technologies in various sectors. With the aim of becoming a benchmark in its field, the DaSCI promotes the transfer of scientific knowledge to the socio-economic environment, thus contributing to technological progress and the digitisation of industry.

Inicio
Figure 1: The proposed method consists of three phases. First, a complete relational network is generated in which the links reflect the similarity between accommodations. Then, the Pathfinder algorithm is applied to eliminate redundant connections and highlight the most relevant relationships. In the second phase, a force-based visualisation algorithm is used to represent the network and community detection techniques are applied. Finally, the visual map is enriched by adding attribute and community information using colour and size scales. In addition, communities are analysed and compared according to their centroids, allowing a better understanding of their characteristics and relationships.

Contacts:

Óscar Cordón García
Departamento de Ciencias de la Computación e Inteligencia Artificial
Escuela Técnica Superior de Ingenierías Informática y de Telecomunicación
Instituto DaSCI
Universidad de Granada
Correo electrónico: ocordon@decsai.ugr.es

Víctor A. Vargas Pérez
Departamento de Ciencias de la Computación e Inteligencia Artificial
Escuela Técnica Superior de Ingenierías Informática y de Telecomunicación
Instituto DaSCI
Universidad de Granada
Correo electrónico: victorvp@ugr.es

Bibliographic reference:

Víctor A. Vargas-Pérez, Oscar Cordón, Manuel Chica, Juan M. Hernández,
Social network of peer-to-peer accommodations for a visual decision support system in tourism: The case of the Canary Islands, Socio-Economic Planning Sciences, Volume 98, 2025, 102145, ISSN 0038-0121, https://doi.org/10.1016/j.seps.2024.102145