Service-oriented computing (SOC) enables organizations and individual users to discover openly-accessible capabilities realized as services over the Internet. However, service registries can potentially be very large preventing organizations from discovering services in real-time. In fact, consumers may not be aware of the services that can be of most benefit to them. In our work, we introduce a web service recommender system that proactively discovers and manages web services. This paper focuses on the underlying search and ranking algorithms that enable the recommendations. As an innovation, we have analyzed real, fully-operational web services currently available on the Internet and, as a result, have discovered insights into how real web service messages are defined. Using these general naming tendencies coupled with enhanced syntactical methods, we are able to aggregate services by their messages and accurately suggest candidate services to users as a part of daily routines.