Research On Methods For Web-based Open Source Intelligence And Analytics

In the age of Big Data,Open Source Intelligence(OSI)analysis becomes more and more important during war and none-war actions.Huge number of data provides a wealth of resources to intelligence analytics,however,the cyberspace of Open Source intelligence have the character of Uncertainty,Diversity,and Complexity(UDC),which forced us to reconsider the theory and technology of intelligence analytics.We proposed an ingelligence analytics method based on artificial societies,computational experiment,parrellel execution(ACP)approach,in order to solve three aspects problem: descriptive analytics,predictive analytics,and prescriptive analytics.The framework based on Popper's "three world" theory,Brooks knowledge and intelligence equation,by using the ACP theoretical approach to constructing an open-source information in an interactive environment intelligence analytics system.The main contributions of this research are as follows:1.Descriptive analytics: We analysis the charactors of intelligence invironment in the perspective of human-information-channel.We constructe a social sensor network to collect web data;propose a SI2 R model to simulated specific information interaction among Sina Microbloggers;propose a latent cyber organizations extract method based on the individuals' sentiment vector.2.Predictive analytics: We predict the development of open source social events based on the computational experiments.An artificial society model is constructed in emerging social media platforms.A scenarios based computational experiment method was proposed.According to open source intelligence for specific domain needs,we conducted experiments in four scenarios,the result shows that many complex experiments can run on artificial society,which can predict the possibility of outcomes when interference by some emergencies and sensitive information.3.Prescriptive analytics: A parallel execution method for open source intelligence inspirement are applied in reality world.On the one hand,the computational experiments results lead us to find the possible strategies for intelligence inspirement.On the other hand,the real world system feed the artificial system to grow more accurately.According to empirical studies,our parallel execution method not only can lead us to specific intelligence needs,but also can bring us some clues for next inspirement.To sum up,we propose an open source intelligence analytics framework,based on it concept and method,we applied it to specific intelligence inspirement.Our research established a theoretical and practical basis for military open source intelligence analytics.