A Novel Intelligent Recommendation Algorithm based on Multidimensional Data Analysis of User Portrait

The collaborative filtering recommendation model needs to adapt to low-quality sparse data, while taking into account high-quality recommendation effects, and fuzzy sets and user portraits based on neural network calculations can solve this problem very well, raising the recommendation quality to a new level. In the designed model, novel intelligent recommendation algorithm based on the multidimensional data analysis of user portrait is studied. In the designed model, the multidimensional data analysis paradigm is first discussed, then, the user portrait analysis model is discussed. The user behavior labels should be properly recombined according to the needs of enterprise management, and the virtual behavior characteristics of ordinary users and leading users should be sorted out. With this additional model, the intelligent recommendation algorithm is implemented with the gradient descent model. In the experiment section, the MovieLens-100K and the MovieLens-1M data sets are used to test the performance, and the accuracy is considered as the standard test. The comparison results show that the proposed algorithm has the better recommendation performance.