Trends in Web APIs Classification, Selection, and Composition

The number of web services and web APIs is still growing. Due to that, the manual selection and possible manual composition of the services is difficult, which is why the research of automatic classification, selection and composition of the services is still actual. This paper reviews recent research on this topic, with the goal of defining the main topics and directions of future research. From the set of 80 read recent conference and journal articles, 30 most relevant papers were singled out. A review of the literature shows that the greatest potential lies in the topic of new services and web APIs (for example, Internet of Things services, machine and deep learning services and APIs) and approaches to classify, select and compose these types of services. The mentioned services have many specific features, and there are still no things, machine or deep learning as a services specific approaches in the existing literature.