A Recommendation Algorithm of Insurance’s Productss Based on Optimal Collaborative Filtering

Abstract:
In recent decades, insurance industry has developed rapidly due to the improvement of people’s living standard and awareness of risk aversion and property safety. However, since 2019, insurance premium growth has slowed down gradually, basically saying goodbye to the era of high growth, the overall development of the insurance industry in the future will change from high-speed growth to high-quality development. To face the current marketing environment and diverse demand of customer, we need to raise the concept about e-commerce and provide customized products to different kinds of customer. In order to solve the problem, a recommendation algorithm of insurance’s products based on optimal Collaborative Filtering is presented. The algorithm using the data of costumer’s basic information and online’s event collected by APP installed on mobile phone to build user-profile about user’s preference of products to alleviate problems of cold start and sparsity in recommendation system. Compare to the traditional CF algorithm, the accuracy of the CF-User Portrait algorithm has significantly improved in the same datasets.