Relation extraction is a fundamental task in information extraction, which is to identify the semanticrelationships between two entities in the text. In this paper, deep belief nets (DBN), which is a classifier of acombination of several unsupervised learning networks, named RBM (restricted Boltzmann machine) and asupervised learning network named BP (back-propagation), is presented to detect and classify the relationshipsamong Chinese name entities. The RBM layers maintain as much information as possible when feature vectors aretransferred to next layer. The BP layer is trained to classify the features generated by the last RBM layer. Theexperiments are conducted on the Automatic Content Extraction 2004 dataset. This paper proves that acharacter-based feature is more suitable for Chinese relation extraction than a word-based feature. In addition, thepaper also performs a set of experiments to assess the Chinese relation extraction on different assumptions of anentity categorization feature. These experiments showed the comparison among models with correct entity types andimperfect entity type classified by DBN and without entity type. The results show that DBN is a successfulapproach in the high-dimensional-feature-space information extraction task. It outperforms state-of-the-art learningmodels such as SVM and back-propagation networks.