This presentation by Ctrip algorithm expert Wei Peng details the rapid development of large language models, their core capabilities, and practical applications in the travel industry, covering content marketing, multimodal interactions, customer service efficiency, model comparisons, fine‑tuning, and inference performance optimization.
This article presents the design and implementation of the Real‑Time User Understanding Service (RTUS) for the Fliggy travel platform, detailing its architecture, multi‑chain data fusion, model and data reuse techniques, and several AI‑driven algorithms for cold‑start interest representation, intent prediction, and destination forecasting, together with extensive offline and online experimental results.
This article presents a comprehensive overview of Alibaba's recommendation algorithm practice for the Feizhu travel platform, covering the evolution of e‑commerce recommendation frameworks, full‑space CVR estimation models such as ESMM, ESM² and HM³, and the specific challenges and solutions for travel‑centric recommendation scenarios.
The article examines how Ctrip shifted from a traditional OTA to a mobile-first travel platform, achieving 500 weekly innovations over three years, expanding its mobile share to over 75%, and rapidly scaling its engineering team to support complex mobile services.
