The identification of credit risk associated with listed companies in the guaranteed network is essential for mitigating risks contagion and sustaining financial system stability. To this end, we investigated the credit risk identifying model by introducing the guaranteed network structure character into the feature variable systems. The listed companies in China had been adopted as the case. More than 60,000 guarantee records of listed companies had been collected and processed: the guaranteed network was formed and analyzed by the algorithm combing the Breadth First Search (BFS) and Interpretive Structural Modeling (ISM); Given the imbalanced data source, the identifying model mixed by PCA input, random sample and RFC model are acquired through comprehensive comparisons. The credit risk recognition rate reached 93%. Summarily, this paper provides an efficient and practical algorithm to construct and analyze the network topology of the guaranteed circle. The recognition model proposed in this paper is highly robust and effective and will assist in improving the credit guarantee approval process of financial institutions by guiding the credit risk governance of listed companies.