Research on artificial intelligence technology promoting the transformation of American intelligence work: optimization of intelligence production process (2)

Artificial intelligence technology promotes the optimization of intelligence production process: artificial intelligence promotes the transformation of intelligence analysis.

Intelligence analysis is at the heart of intelligence work. The U.S. intelligence community believes that artificial intelligence technology is expected to promote the transformation of intelligence analysis. For a long time, intelligence analysis has adopted a human-centered working model. Intelligence analysis is affected by many factors such as the analyst's experience, experience, and education level. At the same time, intelligence analysis scholars believe that there are two main modes of intelligence analysis work: concept-driven analysis and data-driven analysis, both of which are affected by the cognitive ability of analysts. At the same time, intelligence analysis is often accompanied by misleading and deceiving opponents, which makes intelligence analysis more difficult.

In recent years, with the escalation of non-traditional security threats, the demand for intelligence analysis by US policymakers has continued to increase. Decision makers not only want to know the "what" and "why" of a thing, but also want to know the "what will happen in the future" of the thing. In addition, with the frequent occurrence of global emergencies, the improvement of intelligence analysis efficiency is also the top priority of intelligence transformation.

The U.S. intelligence community believes that, first of all, data tools using deep learning technology can infer the phenomena and trends behind the intelligence data flow and judge the causal relationship of things. The intelligence visualization network constructed by artificial intelligence technology helps to dig out the deep context and potential meaning of events. Second, using graph analysis techniques and machine reasoning can effectively circumvent analysts' cognitive biases. Through automatic correlation and automatic screening of data such as intelligence targets (personnel, buildings, organizational networks), it is beneficial to form a grasp of the overall situation and situation. Thirdly, cleaning, clustering, correlating, and describing nodes through artificial intelligence algorithms can effectively filter conflicting intelligence data and effectively resist intelligence deception. The most important thing is that intelligence data processing and auxiliary analysis system through artificial intelligence technology can improve the automation level of intelligence analysis and improve the efficiency of intelligence analysis.