How Yource’s Intelligent System can predict the outcome of court cases by using AI
Monday, February 22, 2021
Yource teaches its system by combining multiple data sources with input from legal experts to help air passengers enforce their rights in case of flight delays or cancellations.
Enforcing air passenger rights can be a difficult task. Over the years, airlines have implemented various strategies and tricks to avoid having to compensate their passengers for cancelled or delayed flights. This is why legal tech scale-ups have made it their responsibility to help passengers enforce their rights in an easy and reliable way. Yource, with the use of its nine European subsidiaries, like Flight-delayed.co.uk, uses an intelligent system to help passengers in the most reliable way possible. “We created a data lake containing flight related data from all around the globe and together with the knowledge from legal professionals as well as data from successful court cases, we ensure that we can very early on assess the exact probability of a claim being successful in court. This helps us to ensure that passengers who experienced a severe flight delay or cancellation receive the compensation or refund they are legally entitled to”, states Mario Wester, CTO of Yource.
Input from legal professionals gives the system more information
The problem that most passengers face when trying to claim compensation or a refund directly with the airline is, that they usually do not have all the information needed to determine the exact circumstances of their flight delay or cancellation. This leads to airlines being able to claim that the reason for the delay or cancellation was out of their control and therefore the affected passengers are not entitled to compensation. Besides gathering flight related data, such as weather conditions as well as airport and aircraft disruptions from all around the world, Yource strives to improve the accuracy of its system’s algorithms by training it with the input of legal professionals. “During the last 10 years, we have gathered data on millions of claims filed with the company by affected passengers. The information contained in those claims as well as the input of our legal partners across Europe regarding these claims, is now used to improve our system and help it to evaluate the chances of success for new claims”, explains Mario Wester.
Additionally, Yource’s artificial intelligence platform is self-learning and training itself by analysing successful court cases that the algorithms will then automatically take into account. “This process of machine learning has recently enabled us to automate the assessment of more than 80% of our claims.” says Mario Wester.
Use of predictive models to assess the probability of a successful claim
As a legal tech scale-up, the focus is to create a database that includes any circumstances that could possibly lead to a flight delay or cancellation, as well as having a multitude of previous court cases and legal information available in order to help as many passengers as possible to successfully enforce their rights. A predictive model is an algorithm being trained with historical data to classify all the available information in order to determine the exact probability of a claim being successful or not. “Simply put, this means that the algorithm discovers trends and patterns in the data we have available for claims that previously were paid out by the airline. For example, if multiple court decisions state that flight cancellations caused by crew strikes fall under the responsibility of the airline, that means that airlines have to compensate customers for these disruptions. Our system will recognise this and assign newly filed claims for flight cancellations due to crew strikes a higher chance of success”, explains Mario Wester. All of this data combined with a competent team of lawyers form nine European countries enables Yource to know exactly in which cases an airline is responsible for a flight delay or cancellation and obliged to compensate its passengers for the flight disruption.