Group hotspot region analysis method for time-space sequence data

Abstract

The invention discloses a group hotspot region analysis method for time-space sequence data. Firstly factors such as speed, distance and the like are considered; group hotspot regions are obtained based on an original position point, but the obtained hotspot regions do not have time correlation, so that the requirements of researchers cannot be well met; and therefore, a method for constructing the hotspot regions with the time correlation based on time division of dynamic adjustment and combination is proposed, namely, when the group hotspot regions are scanned and integrated, each moment is subjected to number-of-people statistics according to a time dimension as a reference dimension, the moments meeting a requirement are recorded in a table, and after the statistics is finished, the current hotpot regions can be divided into the hotspot regions with the time correlation by using linear time. Behavior features of group users can be researched at a deeper level, and related laws can be analyzed. Time-space multidimensional data is subjected to data analysis, so that basic support is provided for privacy protection of spatial big data; and the method has wide application prospects.

Description

technical field [0001] The invention belongs to the field of user group spatio-temporal data clustering feature analysis, and in particular relates to a group hotspot area analysis method for spatio-temporal sequence data. Background technique [0002] Spatial big data is a typical high-dimensional data collection that conforms to spatiotemporal characteristics because it involves multiple information such as user identity, time, and spatial location (such as longitude, latitude, and altitude). In this kind of high-dimensional data integrated in time and space, time, users, and location are closely related. Using traditional high-dimensional data dimensionality reduction algorithms cannot perform dimensionality reduction processing well, and location information includes longitude and latitude. Although there is a certain relationship between altitude characteristics and time, there is no specific measurement index to connect these characteristics, and the research work in r...

AI Technical Summary

Problems solved by technology

However, multi-dimensional user data cannot be directly processed due to the internal relationship between longitude, latitude, altitude, time and other multi-dimensional attributes.

Images