In this paper, we reorganize the nonparallel support vector machines to two types of frameworks. The first type is constructing the hyperplanes separately, where they solve a series of small optimization problems, but it is hard to measure the loss of each sample. The other type is constructing the hyperplanes simultaneously, they solve one big optimization problem with the definite loss of each sample. Based on the framework, we construct a large margin distance-based nonparallel support vector machine for multi-class classification problem, called DNSVM. DNSVM constructs nonparallel hyperplanes with large distance margin by solving an optimization problem. Experimental results on benchmark datasets show the advantages of nonparallel support vector machines, especially our DNSVM.