With the rapid development of the tourism industry, the problem of storing and processing increasingly large amounts of tourism data is receiving increasing attention. To solve the contradiction between users' personalised needs and information overload in tourism services, the improved FP-growth correlation analysis algorithm is used to achieve the recommendation of tourism services of interest to users. Based on the Hadoop cloud computing platform, a recommendation model for tourism services is constructed by fusing cloud computing technology with the improved FP-growth algorithm. The results show that as the number of users increases, the score of the research algorithm increases by about 40% compared to other recommendation algorithms, with a score of about 0.92. The research algorithm is the most effective in filtering out non-interesting travel service items and outdated data from users. When facing new users, the recall of the research algorithm is 0.832, the accuracy is 0.867, and the score of the research algorithm is 0.916. This research is innovative in implementing travel service recommendation on Hadoop cloud computing platform, which not only improves the accuracy of travel service recommendation, but also solves the cold start problem of the recommendation system.