App intelligent recommendation method, device and computer-readable storage medium

Abstract

The invention relates to an artificial intelligence technology, and discloses an APP intelligent recommendation method, comprising: acquiring a user behavior data set, compiling the user behavior data set into an initial data set, and dividing the initial data set into initial indicators The initial index set is de-spaced, filled with missing values, and normalized to obtain a standard index set; the standard index set is analyzed according to the covariance analysis method to obtain a feature index set; the feature index set is and the classification set are input into the pre-built intelligent recommendation model for training; the user's demand instruction for the APP is received, the demand instruction is input into the intelligent recommendation model, and the APP that meets the user's demand is recommended. The present invention also provides an APP intelligent recommendation device and a computer-readable storage medium. The present invention can realize an accurate and efficient APP intelligent recommendation function.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, and in particular, to a method, a device and a computer-readable storage medium for personalizing an APP recommendation based on user behavior records. Background technique [0002] In recent years, with the rapid development of science and technology, a variety of APP (Application) applications have sprung up like mushrooms after a rain. How to choose an APP that meets the needs of users from the numerous APPs has become an urgent problem to be solved. Most of the current technical solutions are decision tree models or naive Bayesian models based on the principles of counting and scoring, recommending the app with the most downloads or the app with the highest rating to other users. Therefore, in today's intelligent era, the recommendation accuracy is not high, and the recommendation method is not intelligent enough to make recommendations based on the actual needs of use...

AI Technical Summary

Problems solved by technology

Most of the current technical solutions are decision tree models or naive Bayesian models based on counting and scoring principles, which recommend the most downloaded APP or the highest rated APP to other users, but because this method does not take into account the differences of different users Actual needs, so in today's intelligent era, the recommendation accuracy rate is not high, the recommendation method is not smart enough, and the recommendation cannot be combined with the actual needs of users

Images