A semantic analysis of polysemy, a case study of LFM

A semantic analysis of polysemy, taking LFM as an example:

A better diagram:

For a given set of user behavior datasets (the dataset contains all user, all of the item, and each user has an item list of behaviors), after using LFM to model it, we can get the models as shown in the following figure: (assuming that the dataset has 3 user, 4 item, The number of classifications for LFM modeling is 4)

The R matrix is the User-item matrix, and the matrix value Rij represents the user I's interest in item J, which is exactly the value we require. For a user, after calculating his interest in all the item, he can sort and make recommendations. The LFM algorithm extracts several topics from the data set as a bridge between the user and item, representing the R matrix as the P-matrix and the Q-matrix. where the P-matrix is the User-class matrix, the matrix value Pij represents the user I's interest in class J, the Q matrix Class-item matrix, and the matrix value Qij represents the weight of item J in Class I, the higher the weight, the more it can be represented by the class.