Progressive multi-dimensional pattern extraction and anomaly detection visual analysis method for spatio-temporal data

Abstract

The invention discloses a progressive multi-dimensional mode extraction and anomaly detection visual analysis method for spatio-temporal data, and relates to the field of spatio-temporal data visualization. The method comprises the following steps of: modeling the multi-dimensional spatio-temporal data into a continuous tensor time sequence, taking a tensor of a latest time period as a current tensor, and performing weighted averaging on data tensors close to a plurality of time periods to obtain a historical tensor; then, respectively carrying out multi-dimensional mode extraction on the historical tensor and the current tensor by using a tensor decomposition method to obtain two groups of rank-one components for describing potential modes in the historical data and the current data; andfinally, calculating a region and a time anomaly score of the current time period based on the difference of the two groups of modes, and visually displaying the difference of the modes in combinationwith a visualization technology to provide support for interpretation of the abnormal modes.

Description

technical field [0001] The invention relates to the field of spatio-temporal data visualization, in particular to a progressive multi-dimensional pattern extraction and anomaly detection visual analysis method for spatio-temporal data. Background technique [0002] With the continuous improvement of urban informatization and the increasing abundance of sensor equipment, a large amount of urban spatio-temporal data is continuously and scientifically collected, enabling the process of human activities to be recorded more comprehensively. These urban data contain a large amount of human behavior information, with multi-dimensional characteristics such as time, space and attributes. In order to effectively mine the valuable information hidden in complex big data and help analysts understand social operations, pattern extraction and anomaly detection for urban spatiotemporal data are crucial. In fields such as social networks, smart medical care, smart transportation, and smart ...

AI Technical Summary

Problems solved by technology

However, there is still a lack of an integrated visual analysis framework that fully supports conventional pattern extraction, anomaly detection, and anomaly interpretation functions, thus limiting analysts to draw more comprehensive and accurate conclusions

In addition, exploring multidimensional patterns in data is a tedious process in most visualization systems, which do not provide effective means to help users quickly and comprehensively understand the patterns, and lack interactive analysis of regular patterns and anomalies in space and time means, making it difficult to explain these anomalies

Images