摘要：

Abstract:

【Purpose/significance】As a complex public opinion field that serves as a controlled and unrestricted information intermedi⁃

ary, there exists a blind spot in the perception of adversarial forces' intelligence, and it is urgent to establish a recognition model to

control it【. Method/process】On the basis of the analysis of the characteristics of the complex public opinion field and the integration of

CTF-MISC, the model building idea is determined and five key implementation steps are designed: confrontation situation determina⁃

tion, opinion theme density analysis, confrontation force situation assessment, opinion theme - confrontation force event pair extraction

and edge computing of the confrontation force subject. Then, the intelligence information summary is refined by giving labels, aggregat⁃

ing and generating portraits, and the application effectiveness is tested through empirical analysis.【Result/conclusion】Empirical evi⁃

dence shows that the model can obtain identification results of three types of complex public opinion fields, Merchant, Resistance, and

Defense, and provide fine-grained feature analysis for opposing forces. The method performance experiment shows that the model asso⁃

ciation method can maintain a leading rate of 13.66% or above, and the effect is good.【Innovation/limitation】The innovation lies in

the model's ability to effectively capture the opposing forces and explain important features, but the limitation is that it fails to fully fo⁃

cus on the humanistic elements such as the historical origins and political biases of the opposing forces, and looks forward to continu⁃

ous improvement.