Node leaf convergence method and device based on tree structure

The invention discloses a node leaf convergence method and device based on a tree structure, and relates to the technical field of big data. A specific embodiment of the method comprises the following steps: splitting a multi-leaf node in a current layer into a plurality of single-leaf nodes; performing column-to-row conversion on each leaf in the single-leaf node set, and performing deduplication on each leaf after column-to-row conversion to obtain a deduplication leaf set; wherein the single-leaf node set comprises the single-leaf nodes in the current layer and the single-leaf nodes split by the multi-leaf nodes; and converging each leaf in the de-duplicated leaf set into a leaf of the node of the upper layer. According to the embodiment, the technical problem that an existing database cannot provide an aggregation method of de-duplicated leaves can be solved. 本发明公开了基于树型结构的节点叶子汇聚方法和装置，涉及大数据技术领域。该方法的一具体实施方式包括：将当前层中的多叶子节点拆分成多个单叶子节点；对单叶子节点集合中的各个叶子进行列转行，并对列转行后的各个叶子进行去重，得到去重叶子集合；其中，所述单叶子节点集合包括所述当前层中的单叶子节点和由所述多叶