Service oriented architecture (SOA) is an architectural paradigm for building enterprise service applications that use services available in a cloud computing environment. In recent days, with the advancement of cloud computing the usage of Web Services has increased rapidly. Therefore, service discovery has become one of the essential challenging factors nowadays for the service consumers. This paper proposes an implementation of dynamic service discovery based on term vector space model using mathematical tool, called MATLAB 7.12. Here, the performance analysis of the proposed model is also describes to support discovering servicers in heterogeneous environments, and show that proposed system performs efficient and scalable in terms of fast searching.