Human group relation network analysis method and device

Abstract

The invention discloses a human group relation network analysis method and device. The method comprises the steps of collecting video data of multiple sites based on a set data collection format, wherein the collected data comprises human face characteristic value and the structural information of the human face characteristic value; conducting the time delayed relation filtering, geological location mapping, behavioral mode judgment and weight relation network construction on the collected data to obtain the basic information of the human group relation network, conducting relation views filtering to the basic information of the relation network to implement a flexible and configurable relation rendering. The invention further discloses a corresponding device. The video image collection and the human face recognition technologies are introduced to the large scope human group relation network analysis, the broadness and unity of the data sources are achieved, and the broad and accurate human group relation analysis can be achieved.

Description

technical field [0001] The invention relates to the field of data processing, in particular to a method and device for analyzing a crowd relationship network. Background technique [0002] In the field of data intelligent analysis and mining, it is often through in-depth analysis and processing of massive data to discover useful patterns hidden in the data, so as to solve various problem scenarios in daily social life. Such as smart city / security, business intelligence (English: Business Intelligence, referred to as: BI) and other application scenarios, all have strong interests and application pain points in human activities, especially related models of crowd (social) activities. These models include Specific activity time patterns, location patterns, trajectory patterns of the crowd and the hidden commercial value behind these patterns. But in these models, the crowd relationship network is a relatively high-value and deep-level model information, which contains high com...

AI Technical Summary

Problems solved by technology

However, these fields still have the following shortcomings: data usually depends on the accuracy of human input, but people usually use some unreal identity information in these activities; identity information based on accounts and other information is easy to be used by multiple parties , leading to a lot of noise data, affecting the accuracy of the analysis

[0006] To sum up the above problems, through the current conventional information sources and analysis methods, first, it is impossible to obtain a solution for the analysis of a wide range of people’s relationship networks. have a certain impact

Images