The problem of evaluating node importance in complex networks has been an active area of research in recent years, and many algorithms and software tools were developed. Most of the current algorithms deal with single criterion ranking, such as degree of nodes. However, in some real world networks because of their large-scale nodes and complex relationship, it is crucial to incorporate multi-criteria ranking into node importance evaluation methods. We proposed a unify multiple metrics evaluating framework of node importance with non-conflict equivalent class. We summarized five rules based on intuitions to measure node importance. Under these five rules we employ the equivalence classes of partial order relationship to rank the node importance. We evaluated our algorithm with three real world typical networks, which are the metabolic network, the dolphins network and the Zachary karate club network. Results show that the multi-criteria ranking approach can be used to evaluate the importance of nodes in real-world networks. We further compared our method with other traditional algorithms such as Page Rank. Our results demonstrate that applying multiple metric into node importance evaluation can make the ranking result more reasonable in some real-world applications.