针对基于概念格对数据的规则挖掘中,概念格结构的复杂度随着形式背景的复杂化呈指数递增的问题,本文从树图的角度研究决策形式背景的属性约简问题,以树型结构为基础,首先提出了新的强弱背景的判定方法;进一步定义了条件树对象集与决策外延的相关函数,并以此作为启发信息,设计了基于树图的逐层属性约简算法,避免了计算区分矩阵这个既消耗时间又消耗空间的过程;最后在属性约简树的基础上,给出了约简后的概念树生成算法,以简洁、直观的概念树的代替错综复杂的概念格.理论分析和实验结果表明,该算法是有效可行的. In the rule mining of data based on concept lattice, the complexity of concept lattice structure increases exponentially with the complexity of formal context. This paper studies the problem of attribute reduction of decision-making formal context from the perspective of tree graph, , A new method to determine the background of strong and weak is first proposed. The related function of the set of condition tree and decision-making extension is further defined, and as a heuristic information, a tree-based layer-by-layer attribute reduction algorithm is designed to avoid Finally, based on the attribute reduction tree, a reduced concept tree generation algorithm is given to replace the intricate concept lattice with a concise and intuitive concept tree. Theoretical analysis and experimental results show that the algorithm is effective and feasible.