可树形分解的联想过程分析

将联想思维过程抽象成联想图，该图是以变量之间的相关性和条件相关性表示的正态分布，进而讨论了联想图的可树型分解问题，包括可树型分解条件及其树重构算法，可树形分解的联想思维不仅是发散的，而且是递进的，因而具有更高的创造性。这些结论不仅可用于评价人类的联想思维过程，而且可通过对联想思维过程重构来刺激创造性思维的产生。

This paper models the reminding process as a reminding graph following the normal distribution represented by the dependence and conditional dependence of variables, and then discusses its properties that are much coherent with our intuition. Particularly, the paper copes with the problems about the thinking with the treedecomposable reminding graph, including its decomposing condition and reconstructing algorithms. The results can be used to evaluate the human reminding processes as well as to help human produce the creative ideas through reconstructing the human reminding processes.