Mashup refers to a sort of web application developed by reusing or combining web API services, which are very popular software components for building distributed applications. As the number of open web APIs increases, to find suitable web APIs for Mashup creation, however, becomes a challenging issue. To address this issue, a number of web API service recommendation methods have been proposed. Content-based methods rely on the description of the service candidates and the user's request to make recommendations. Collaborative filtering-based methods use the invocation records of services generated by a set of users to make recommendations. There are also some studies employing both the description and invocation records of services to make recommendations. In this paper, we survey the state-of-the-art web API service recommendation methods, and discuss their characteristics and differences. We also present some possible future research directions in this paper.
展开▼