Relationship defining according to the ideflx method , on the basis of building a relationship matrix of basic information entities , seven information entities were validated , including customer entity , member entity , project entity , mold entity , assignment entity , member of assignment entity , and file entity . and then , relationships between these information entities were discussed ; 3 （ 2 ）定义联系按照idefix方法，通过建立实体联系矩阵并进行联系分析，最终确定信息实体包括客户信息、成员信息、项目信息、模具信息、任务信息、任务成员和文件信息七大项，并建立实体联系定义表，构造出实体级图。
However , it suffers from great complexity and low efficiency . the hash - tree - based approach is a slight modification of the apriori algorithm , which , again , we find not satisfying . to further improve the efficiency , we devise another approach based on a different yet similar tree structure 第一种算法是从混合规则的定义出发采用直接的方法生成大项集，算法的复杂度较大，效率不高；第二种算法是对apriori算法稍加修改得到的，也不够令人满意；为进一步提高执行效率，我们设计了基于一种树结构的算法。
