Traveling and discovering places is on almost everyone’s bucket list. In theory, most of us should feel excitement (or at the very least calm) regarding our travels. In reality, however, there is much more to it than hopping on a plane and arriving at the destination.
Between searching for travel destinations, paying for bookings, organizing schedules, and digging for a city’s hidden gems, what should be exciting can quickly become incredibly stressful.
A clear overview of a trip before the journey begins can be a genuine remedy. This is even more important for business trips, where a well-arranged agenda is key to prevent losing valuable time. So how can you handle this organizational challenge and reduce the efforts of planning your next trip?
TripIt for Travel Planning
Travel-organizing app TripIt from Concur provides a complete solution to coordinate your travel schedule. From the moment you leave your home to the moment you return, TripIt acts as a personal travel assistant, reminding you of travel times, updates, addresses, and other important things you need to be aware of while traveling.
TripIt from Concur is your personal travel assistant
So how does TripIt work and what’s the technology behind it? Most of us have experienced the stress of having to find a flight number in an email buried somewhere among hundreds, all while trying to get to our flight on time. For TripIt users, clicking through multiple apps and scouring emails to find the travel information you need will no longer be an issue. TripIt relies on a set of rules that are part of its Datamapper automated parsing engine technology to capture relevant data from the traveler’s shared confirmation emails. In short, the app extracts flight schedules or hotel stays from the provided emails and instantly organizes them into one easy-to-read master itinerary. The application promotes the confidential handling of personally identifiable information.
Machine Learning for Better Travel
Like many other enterprises, TripIt currently plans to integrate automated machine learning methods. The new technology, which makes machines or programs smarter without being explicitly programmed, helps to improve performance and user experience in several industries. In the travel industry, it can automate tedious manual processes for travel agencies and related service providers while making travel planning easy for the end-consumer.
Before introducing machine learning into Datamapper, TripIt relied on a set of manually entered rules, also known as Datamaps, that were tailored to the various email templates used by the travel vendors. The human-powered system was labor-intensive in terms of generation of templates and maintenance. In an environment that is comprised of thousands of travel providers — airline agencies, hotel bookings, etc. — that often change their templates and employ multiple models based on the user’s profile, there was an acute need to build an automated system.
In a current project collaboration with the SAP Machine Learning Research team and Tel Aviv University (TAU), TripIt works on integrating machine learning technology into its Datamapper. The research team collaborates with top-tier universities to solve a wide domain of complex machine learning problems, building a bridge between first-class machine learning research and SAP’s enterprise solutions.
Academic Support to Integrate Machine Learning
In the first accomplished milestone of the project, a deep learning-powered text classifier was developed to automatically classify itineraries, including air travel, hotel, and car service, as well as to identify the vendor. The underlying technique applied during the research collaboration falls under the deep learning concept of natural language processing (NLP), which is a sub concept of machine learning. As such, NLP is useful for extracting information from a wide-range of domains and formats — from PDF documents over social media to emails.
According to Dr. Tassilo Klein, senior researcher for Machine Learning at SAP, “NLP has been experiencing unprecedented boost in performance with the recent progress in deep learning technology. It was therefore only logical to launch initiatives in this domain for the first attempts to make enterprise applications smart.”
TripIt has grown to more than 14 million users and continues to work on refinement. Upon conclusion of the project, SAP and TAU plan to deliver a deep learning solution that enables TripIt to automatically identify and extract itinerary data, reducing the need to manually create and maintain templates. Yet another example of machine learning in the business sphere, the project further supports TripIt to stay true to its slogan: “You handle the booking. We take care of the rest.”
Find out more about TripIt and download the app at www.tripit.com.