Language is an enormous barrier to global e-commerce. Internet users favor visiting or shopping on Web sites presented in their native language. This research proposes a shopbot with a multilingual ontology to overcome this language barrier. The shopbot, called WebShopper, collects product data from online vendors over the Web and enables customers to execute semantic search using different languages. An empirical study is conducted to evaluate this shopbot. The result shows that customers are able to reach more products and find the real bargains using WebShopper, and the proposed semantic search can improve search results and user satisfaction.