Aviation service customization system and method based on federated learning technology platform

Abstract

The invention discloses an aviation service customization system and method based on a federated learning technology platform. The system comprises a federated learning technology platform, a data source module, a data processing module, a local training module and a cloud module. The method comprises the following steps of establishing own local models for airlines and partners by utilizing own source data respectively; and then uploading parameters and gradients of the model to a federated learning technology platform, realizing fusion on the federated learning technology platform so as to establish a new virtual model, performing prediction and updating by utilizing the virtual model, and finally establishing a federated personalized recommendation model. The model established through the method can provide personalized and accurate product recommendation for aviation customers, meanwhile, the modeling process is achieved through a federated learning technology platform, and safetyand privacy of data are guaranteed.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to an aviation service customization system and method based on a federated learning technology platform. Background technique [0002] In the increasingly competitive civil aviation industry, in addition to the profits obtained from ticket sales, the impact of additional income on airline profitability is expanding. However, under the original marketing management model, airline customers are in a state of passive management, and the sales of airline products and services are not in the right way, lacking marketing precision, and it is difficult to meet the requirements of market changes. [0003] With the further development of big data, the use of big data technology to form a "prediction" effect through long-term storage and correlation analysis of massive amounts of time, which is helpful for the precise mining of consumers' potential consumption needs, passenger value analys...

AI Technical Summary

Problems solved by technology

However, there are difficulties in establishing an accurate personalized service recommendation model for airlines: 1. There is a lack of accurate user portraits to reflect the travel needs or preferences of passengers; 2. Passenger behavior data is usually very sparse; For most people, air travel is usually for some specific needs, such as business or vacation, rather than daily travel; 4. The travel frequency of passengers and the needs of different routes are in a long-tail distribution

Therefore, most passenger behavior data are overwhelmed by low-frequency travelers and cannot be modeled well

Images