In this paper, we focus on the domain ontology acquisition from Chinese corpus by extracting rules designed for Chinese phrases. These rules are noun sequences with part-of-speech tags. Experiments show that this process can construct domain ontology prototypes efficiently and effectively.