而PCRW根据节点的邻域对路径进行评估，考虑到每一步节点类型选择对关联

PCRW evaluates the path according to the neighborhood of the node, and considering the influence of node type selection at each step on the degree of relevance, a probability selection model is added to each step of the walk, which contributes to the final result more in line with the actual situation. <br><br>However, the above two similarity measurement methods lack consideration of node type label selection, and the meta-path must be given in advance.

PCRW evaluates the path according to the node's neighborhood, taking into account the influence of each node type selection on correlation degree, and adds a probability selection model to each step of the walk, contributing the final results that are more in line with the actual situation.<br><br>However, both of these similarity measures lack consideration for node type label selection and must be given a metapath in advance.

Pcrw evaluates the path according to the neighborhood of the node. Considering the influence of node type selection on the correlation degree at each step, pcrw adds a probability selection model to each step, which contributes to the final result more in line with the actual situation.<br>However, the above two similarity measurement methods lack the consideration of node type label selection, and must give the meta path in advance.<br>