Home >Open Calls >Experiments > DATA-DRIVEN HOTEL

Deployment of Data Use Cases to Provide a Hotel with Predictive and Data-sharing Capacity



Vast majority of hotels currently rely on monolithic information management systems without the necessary operating systems and procedures to make data-driven decisions.
AdQuiver's experiment aims to transform this approach by helping hotels move from descriptive to predictive data exploitation.
The experiment will focus on developing three use cases which will enable hotels to predict booking and incomes, cancellations, and cluster demand and will use relevant data from the Hotel transactional systems, digital ads ecosystem, and EUHubs4Data datasets.
This advanced analytics approach will apply supervised and unsupervised machine learning models that go beyond traditional data analysis to yield the greatest benefit to the hotel and enable them to make more informed decisions that improve their daily operations and generate competitive intelligence.

Main Objectives

The main objectives of the project:

  • Enable data-driven decision-making in a Hotel chain by transforming its data management approach from descriptive analytics to predictive and advanced analytics.
  • Provide the hotel with three use cases, including a dynamic booking and income predictor, probabilistic cancellation classifier, and unsupervised demand clustering.
  • Increase efficiency in resource allocation, reduce financial risk, and personalize offers and anticipate cancellations through three use cases.
  • Provide a solution which generates valuable insights for all functional areas of the hotel and offers near-real-time dashboards to accurately predict KPI behaviour for better decision-making.
  • Enable data-driven decision-making to become more common in the tourism sector's digital transition, allowing hotels to address challenges such as predictive planning, resource efficiency, and personalised service.
Main innovations

This experiment involves experimental development activities (TRLs 7-9) to extend the use of data management, AI, and advanced analytics in a hotel company. Prediction algorithms make use cases available for anticipation in decision-making through data visualization tools. Technological elements facilitating the innovations include data infrastructure (Data Collector, Data Warehouse among others), BI, and advanced analytics using AI, ML, and deep learning techniques.


The hotel industry will anticipate relevant industry KPIs that will enable to optimise resource allocation which may have a direct impact on the environment, reduces financial risk among other benefits.
The implementation of the experiment will also positively contribute to creating qualified employment in the Hotel industry resulting from digitalisation and improved competitiveness in the sector owing to the impact of this innovation.

Key Exploitable Results (KERS)

AdQuiver will conceptualise, develop and put into exploitation the following three use cases that will be made available to clients through Dashboard as a Service:
1 - Dashboard for predicting the number of reservations and income
2 - Cancellation prediction dashboard
3 - Dashboard for customer clustering at the destination, performing predictive planning of demand.

Technical KPIS

• Milestone Definition: Unified and pseudonymized datasets in CSV format hosted on our Data Lake, so that it can be accessible by the Analytics platforms like EGI (Notebooks) service
• Unit of measurement: Number of datasets in CSV format hosted on our Data Lake
• Expected value at Month 5: (4)
• Expected value at Month 10: (6)
• Milestone Definition: Predictive model with a minimum value of precision
• Unit of measurement: Classification use case, Cancellation prediction (% accuracy) / Regression Use Case, Prediction of reservations and income (RMSE) / Clusterization of clients (Number of Clusters)
• Expected value at Month 5: Accuracy min 70% | less than 1,5 rmse | 1-3 clusters)
• Expected value at Month 10: Greater than or equal to 75% | less than 1 RMSE | 3-5 clusters)
• Milestone Definition: Dashboards developed with the visualization of the results of the predictive models that have achieved the minimum established precision to
date, and which will be made available to the end user through the AdQuiver website.
• Unit of measurement: Number of dashboards
• Expected value at Month 5: (1)
• Expected value at Month 10: (3)
M 4
• Milestone Definition: Assistance to a EUHubs4Data event to present and promote the experiment and the beneficiary company. (Fixed and Mandatory)
• Unit of measurement: Number of events
• Expected value at Month 5: (0)
• Expected value at Month 10: (1)
• Milestone Definition: Datasets with pseudonymized data in parquet, avro and csv format, with the results of the predictive models, accompanied by an ontology for each
• Unit of measurement: Number of datasets
• Expected value at Month 5: (1)
• Expected value at Month 10: (4)


Adquiver is a tech company specialised in the Travel sector that offers Data-Driven solutions. MoirAI, its Big Data Platform integrates multiple data sources from the digital ecosystem and relevant transactional systems to provide Advance Data & Analytics and Media services.