Mobility prediction method, system and device based on user classification

Abstract

The invention belongs to the field of wireless communication and data mining, particularly relates to a mobility prediction method, system and device based on user classification, and aims at solvingthe problems that an existing user mobility detection method is poor in prediction accuracy and low in precision. The method comprises the following steps: constructing a historical movement track sequence of a to-be-predicted user as a first sequence; based on the first sequence, obtaining a moving track sequence in a set time period as a second sequence, and obtaining a user type through a preset user type classification rule; obtaining the maximum step length k based on the user type, and constructing Markov state transition probability matrixes from 1 to k; obtaining the prediction accuracy of transferring to the next position from each position in the second sequence based on each matrix, and calculating the weight of the Markov model in each step; and calculating the probability of reaching each selected candidate position through a weighted Markov model, and taking the candidate position with the maximum probability as the next prediction position. According to the invention, the prediction accuracy and precision are improved.

Description

technical field [0001] The invention belongs to the fields of wireless communication and data mining, and in particular relates to a mobility prediction method, system and device based on user classification. Background technique [0002] In the mass mobile communication data, there is a lot of valuable information. Based on this information, service providers can design better operation solutions and improve the experience of mobile users. For example, using the user's location information and service preference information contained in mobile communication data, the user's mobile behavior model in time and space can be established, and then the user's behavior pattern can be predicted. Effective trajectory prediction enables service providers to predict user needs in advance, thereby optimizing network resources and reducing network congestion. Users can get the required information faster and get better service. [0003] In addition, the user's trajectory prediction res...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the above-mentioned problems in the prior art, that is, in order to solve the problem that the existing user mobility detection method does not consider the mobility patterns of different users, resulting in poor prediction accuracy and low precision, the first aspect of the present invention proposes A mobility prediction method based on user classification, the method comprising:

Images