CONCLUSION
We advocate future research that aims, on the one hand, to improve dealing with these impediments and, on the other hand, expands the scope of what we can automatically learn about web APIs, for example about required and returned data, response codes, or security mechanisms.