以关联语义链网络为基础，基于大规模Web资源词汇数目有限性及知识的层次性，提出事件及领域背景知识的"全局知识属性-概念知识属性-词汇知识属性"三层关联语义链网络(TALN)模型及构造方法，以避免对大规模Web资源的直接计算。基于事件TALN的网络拓扑结构及知识属性的不确定性、可信度与热点度等参数，计算事件微观知识属性承受的内部力；基于领域TALN及被激活背景知识属性的参数，计算知识属性承受的外部力。然后，探讨这两种力对事件微观知识属性流动和演化过程共同作用的机理。继而，基于信息熵与完美无标度网络，探讨事件知识宏观流动和演化过程的无序性，度量事件知识整体流动与演化过程的动力；基于相邻时刻事件TALN中拓扑结构的变化，获取事件知识宏观流动与演化速度、加速度及方向向量。最后，把上述研究应用到新型语义搜索和最具流动与演化潜力事件和资源的搜索与推荐中；以缓解资源大规模性与人认知能力有限性之间的矛盾。

Based on the Association linked Network (ALN), this project proposes a Three level Association semantic Link Network (TALN) for a web event or a domain, including general knowledge attribute level, concept knowledge attribute level and word knowledge attribute level. On the one hand, the TALN can reveal and express the hierarchy of the knowledge. On the other hand, its scale can be kept not increasing with the large scale web resources because of the limited number of keywords in a domain, and this relative small and stable scale will significantly reduce the computational complexity when meeting a large scale data. An internal force to push the evolution of micro-knowledge attribute is defined and measured through the topology of TALN, uncertainty, credibility, hot degree, and so on; an external force is also defined and measured based on the interaction between TALN and domain knowledge. Then, the combination of two forces, internal one and external one, is investigated for studying the evolution of macro-knowledge attributes. The disorder of macro-knowledge attributes is measured based on information entropy and perfect power law distribution as the force of pushing the evolution of macro-knowledge attributes. After that, the velocity, acceleration and direction are measured by the comparison between the TALN at different times on different or same levels. At last, our proposed models and theories will be applied to a promising application scenario, web events or resources searching by their flow and evolution potentials.