Dietary Sequence Recommendations Based on TransFM

The rapid development of the Internet has changed people's lifestyles, and more and more people choose to eat online. The online catering platform for users inevitably needs to provide personalized recommendations for users, followed by fewer articles on dietary recommendations. This paper proposes a combination of GBDT and TransFM for dietary sequence recommendation. The hybrid model uses GBDT to process high-dimensional features of food data and capture high-order feature interactions, and then uses TransFM to combine the content features of food and mine the short-term interests of users. Perform sequence recommendations. After experimental comparison, the model is superior to the general sequence recommendation model, and has achieved good results on multiple performance indicators.