Research
Global harnesses the explosion of digital data and computational power to better understand travelers’ behaviors and create smarter travel experiences. The dedicated researchers and data scientists of our Research labs are constantly looking for new ways to improve the travel experience through Data and Artificial Intelligence (AI).
Research Focus
The goal of our Research teams is to understand travelers’ behaviors and create a smarter, seamless and more sustainable travel experience. We support the Global Innovation mission of improving the traveler experience before, during and after the trip, and at destination.
"At Global we invest seriously on advancing research on AI & emerging technologies with the ambition to make travel a better, seamless and more sustainable experience."
Research use cases
Global has proven expertise in Operations Research and has been working with Artificial Intelligence in travel since 2011. A.I.-powered solutions are now found at every step of the traveler journey.
Before the trip
- Reinforcement learning for Revenue Management
- Dynamic pricing for airlines
- Choice modeling to understand and predict traveler choices
- Recommender systems for Hotels and Ancillaries
- Price prediction
- Airlines fare family design
- Click-through-rate optimization for Digital Advertising
We use machine learning in flight search to segment travelers and deliver the most relevant results to each traveler group. This has boosted booking by 5-20%.
During the trip
- Recommender systems for transfers
- Baggage weight prediction
- Flight delay prediction
- Disruption management
- Airport optimal gate assignment
- Airport takeoff sequencing optimization
- Seat allocation optimization
Global uses machine learning to optimize flight sequencing. This technology has reduced for a European airport runway time by 50 %, limiting fuel wastage as a result, and inbound delays have been improved by 24% compared to outbound flight delays.
Creating data-driven solutions through research
Our research contributes to the virtuous circle of data. Our Research labs focus on identifying and validating the technical feasibility of delivering new and impactful solutions that consume, generate or enrich data. We use data science and emerging technologies to explore smarter ways for apps to consume data.
We also explore non-AI emerging technology like Internet of Things (IoT), Augmented and Virtual Reality, Blockchain and Robotics to discover new ways to improve the traveler journey and unlock new potential use cases. Data sets generated by these new technologies can contribute to the data-driven virtuous cycle.
Our teams are exploring Natural Language Processing (NLP), Deep Learning and Generative Adversarial Networks (GANs) and use the latest cloud tools to continuously test new algorithms.
Global AI APIs
The Global Research teams have joined forces with Global for Developers to bring a game-changing solution for innovators in travel: Global Artificial Intelligence APIs.
Global Artificial Intelligence APIs give startups and innovators the powerful predictive capabilities of machine-learning models trained with Global data, via API.
Innovators can access these predictive capabilities via easy-to-use REST/JSON APIs on the Amadeus for Developers portal. Then, any developer can start testing the APIs in the Self-Service API catalog. They are accompanied by documentation, code samples, guides and more to get started quickly.
Flight Choice Prediction API
This API uses predictive machine learning models to analyze flight search results and assign each flight a probability of being chosen by the user. Retailers can then identify the most attractive flight deals in the search results and highlight them to the user, creating a more optimized shopping experience.
Trip Purpose Prediction API
The Trip Purpose Prediction API predicts whether a given flight itinerary is most likely for business or leisure travel by analyzing the cities of departure and arrival, the flight dates and the search dates and comparing them against our historical data. This lets you understand the motivations behind a travelers’ searches and offer them the most relevant content.
Flight Delay Prediction
The Global Flight Delay Prediction API uses a machine learning model trained on Global data and a variety of variables like departure and arrival times, dates, airports, carrier or aircraft type to calculate the probability that a flight will be delayed by under 30 minutes, 30-60 minutes, 60-120 minutes or over 120 minutes.
AI-Generated Photos
The AI-Generated Photo API is an experimental project by the Global A.I. Lab that uses the Nvidia Style GAN framework to return a picture of a beach or mountain landscape created by Artificial Intelligence. This image can then be used to enrich your offer without the hassle of sourcing copyrighted images.
Airport On-Time Performance
The Global Airport On-Time Performance API uses a machine learning model trained on Global’ historical data to estimate a percentage of total on-time flight departures for a given airport and date.
Travel Recommendations
The Travel Recommendations API provides travelers personalized destination recommendations based on their location and an input destination, such as a previously searched flight destination or city of interest. The API uses Artificial Intelligence trained on Global historical flight search data to determine which destinations are also popular among travelers with similar profiles, and provides a list of recommended destinations with name, IATA code, coordinates and similarity score.