Currently, the development of tourism big data and artificial intelligence cross-technology provides an important role for the tourism industry to predict and analyze consumer behavior and help the development of the tourism industry. This paper discusses the relationship between tourism big data and tourism consumer behavior, and as a basis for consumer behavior, puts forward the ARMA model and LSTM model based on consumer behavior data and then combines them to design a single-layer LSTM-based combination prediction model. Finally, a famous tourist attraction in GD province is used as an empirical research object, and the model proposed in this paper is used to predict and analyze consumer behavior. The analysis of tourism shopping goods revealed that 44.97% of consumers preferred to buy tourism souvenirs, indicating that the cultural and creative products in Area A are more appealing to tourists.