A growing number of companies are using web services to make their expertise and data available through the network. The current problem is that the content of these web services remains inaccessible to machine processing. Only humans can interpret their contents. The Semantic Web is the new vision of the Web that promises to overcome this difficulty. The concept of semantic web services, is the result of the convergence of the field of web services with the semantic web, indeed its objective is to automate the discovery, selection and composition of web services. In this work, we are interested in the semantic discovery of Web services. The main problem is to automate the discovery of web services to respond to a request from a client. In this sense, firstly we present a conceptual framework and architecture to carry out our approach. The originality of the proposed solution lies in the use of mixed technical tools ranging from semantic models to multi-agents systems, including Matchmaking algorithms. Afterwards, we implement our proposed architecture. To validate our work, we conduct tests with a variety of user queries and a panel of Web services. As part of a case study we consider an online travel organization problem. This problem is a typical web services discovery scenario to apply the concepts of our approach.