Design and Implementation of Insurance Product Recommendation System
JiaPeng Zhang · Frontiers in Computing and Intelligent Systems · 2022
In this paper, a hybrid collaborative filtering recommendation algorithm with Latent Dirichlet Allocation (LDA) and Alternating Least Squares (ALS) is proposed, which can effectively alleviate the cold start problem in traditional recommendation system and improve the accuracy of recommendation. At the same time, real-time recommendation according to the real-time behavior of users is realized in the real-time recommendation system, which effectively improves effectiveness of the recommendation results. In the system construction, user historical data is collected and analyzed by big data tools. Based on the Spark framework, the offline and real-time recommendation of insurance products are effectively combined and realized, with personalized recommendation as an important function. The designed system can improve the purchase rate and experience of users.