Page 165 - Journal of Library Science in China, Vol.45, 2019

164
164 Journal of Library Science in China, Vol.11, 2019

CIA models have shown the following characteristics and trends in recent years. First, the CIA
model is no longer limited to simple descriptive statistics modes, but tends to be a kind of more
comprehensive or combined mathematical mode. Secondly, the construction of the CIA model is
transformed from a static model to a complicated dynamic model, that is, variables changing over
time and their relationships are introduced into the model. Thirdly, the construction and application
of CIA models are based on massive data or big data. Fourthly, the model construction’s focuses
shift from evaluating past achievements to predicting and interpreting future outcomes and
scientific discoveries.
The CIA framework mainly refers to the technical framework for intelligence analysis, which
consists of abstract components and messaging methods for the components’ instances. The CIA
framework can be considered as a constrained support structure designed to solve the open-ended
problem in the process of intelligence analysis. A CIA system is an implementation of the CIA
framework. In recent years, the development and research of CIA framework and system have
shown the following characteristics and trends. First, the analytical framework and system are
evolving from a single architecture to a distributed, cloud computing architecture. Such framework
and system are more in line with the requirements of storage, computing, analysis of multi-source,
heterogeneous, massive data. Secondly, the CIA framework and system have changed from a
traditional relatively closed, dedicated mode to an open, military-civilian integration mode. Thirdly,
the intelligence analysis processes in the framework or system can be customized and automated.
Fourthly, the CIA framework and system are becoming more intelligent with the application of
artificial intelligence technologies.
The CIA algorithm is a series of computation steps and instructions formed under specific
intelligence analysis tasks or intelligence analysis scenarios by the guidance of intelligence
thinking. The CIA algorithm is the prerequisites for automated intelligence analysis so that the
intelligence analysis model can be calculated by machine and finally obtains the computational
results. The CIA algorithm has recently shown some characteristics and trends. First, the explaining
ability of intelligence analysis algorithm is being initiated. Secondly, the open source intelligence-
oriented perception algorithm is increasingly being valued. Thirdly, the algorithm for the fusion of
human analyzed results and computer analyzed results has got more attention.
The CIA is an extension of traditional intelligence analysis and now it becomes one of the
important directions of future development of information science. The CIA will gradually develop
into a new discipline that integrates computational science, social science, cognitive science,
complex science, biochemical science, engineering, and many other academic subjects. For
researchers, opportunities and challenges will coexist. This requires us to meet the challenges of
the future with more innovative, open and bold ideas.