由于传统软件识别矛盾问题的能力不足，限制了计算机在解决矛盾问题方面所发挥的能动作用。可以尝试利用Agent的自主性、社会性和反应性等特性，协助识别矛盾问题。针对Agent缺乏界定矛盾问题核问题的能力的缺点，本文提出一种方案，以可拓信息知识策略形式化体系作为Agent的知识链，利用Agent的通信能力与反应性进行人机交互，引导用户输入基元信息，减轻让Agent识别以自然语言描述的矛盾问题信息的困难，使矛盾问题核问题不断清晰，直到可以进行可拓变换解决矛盾问题。创新点在于结合了可拓学、Agent、本体等几种不同学科领域技术的优势，使得Agent有能力协助人们分析和解决矛盾问题。

Since traditional software lacks the ability of recognizing the contradiction problems, the computer’s functions to solve the contradiction problems are limited. The agents’ autonomy, social ability, and reaction can be used to help recognize the contradiction problems. In view of the drawbacks that agents lack the identification ability for the kernel problem of the contradiction problems, the paper puts forward a scheme. It takes the informationknowledgestrategy formal system as the knowledge chain of agents. It uses the communication ability and reactivity of agents to introduce the users’ basicelement information so that they can reduce the difficulties of understanding the contradiction problem information. The scheme enables the kernel problem of the contradiction problems clearer and clearer until the contradiction problems can be solved by extension transformation. The new idea is combining the advantages of several different domain technologies: extenics, agents and ontology so that agents are able to analyze and solve the contradiction problems.