Bold-and-unconstrained style and Graceful-and-restrained style can characterize poetry's taste, which usually is judged personally, so the assessment is always subjective. If the methods of Machine Learning can be used to assess poetry style, it will be more objective. This paper brings forward a method based on Support Vector Machine (SVM for short) to differentiate bold-and-unconstrained style from graceful-and-restrained style of poetry. In this work, a piece of poetry is expressed using Vector Space Model (VSM for short) first, and then we use information gain to select the poetry's feature terms. At last, we use an SVM-based method to divide the style of poetry. Meanwhile, feature numbers and feature items for poetry style's influence are also analyzed. The performance of the proposed method has been evaluated by a series of experiments with interesting results.