Latent variable model-based user preference extraction method

Abstract

The invention discloses a latent variable model-based user preference extraction method. The method comprises the steps of firstly selecting N commodity relative attributes to form a commodity property set, building according to historical data to obtain a bayesian network, searching in the bayesian network to obtain a maximal semi-clique, and then inserting a latent variable L, showing user preference, into the maximal semi-clique, so as to obtain a latent variable model, wherein L being equal to 1 shows that a user prefers, and L being equal to 0 shows that the user does not prefer; performing parameter learning on the latent variable model to obtain a conditional probability table of various nodes in the latent variable mode; then according to the conditional probability table of the latent variable L, performing user preference extraction: searching to obtain an attribute combination item corresponding to a conditional probability maximum when L is equal to 1, wherein the attribute combination item corresponds to commodity types most preferred by the user; searching to obtain an attribute combination item corresponding to the conditional probability maximum when L is equal to 0, wherein the attribute combination item corresponds to commodity types least preferred by the user. By aiming at the user preference hidden in commodity evaluation data, the more objective and realistic user preference results are extracted by the structure of the bayesian network.

Description

technical field [0001] The invention belongs to the technical field of service information push, and more specifically, relates to a user preference extraction method based on a latent variable model. Background technique [0002] With the continuous development of Web2.0 and the continuous popularization of social media, a large number of users' evaluation data on commodities (commodity evaluation data for short) have been generated in e-commerce applications. To a certain extent, it reflects user preferences and hobbies. In the product evaluation data, the directly observable part includes the attributes of the product itself (such as product type, price, favorable rate), and the user's rating of the product. At the same time, user preferences that cannot be directly observed determine, on the one hand, users’ browsing, purchasing, and evaluation behaviors for products; on the other hand, these behaviors of users also reflect user preferences, that is, users’ preferences ...

AI Technical Summary

Problems solved by technology

However, the above methods pay less attention to practical applications, and do not give hidden variables the actual meaning of specific applications.

In Zhang Fuzheng's paper "Understanding User Behavior Based on Large-Scale Location and Consumption Data", the Bayesian network is used as the basis to express the time series behavior of users as observable variables, and to express the user's novelty-seeking psychological characteristics. and utility preference are represented as latent variables, and the user's behavior pattern is represented through the model generation process, but this method does not involve extracting user preference based on commodity evaluation data

Images