Learning to Recommend Third-Party Library Migration Opportunities at the\n API Level

The manual migration between different third-party libraries represents a\nchallenge for software developers. Developers typically need to explore both\nlibraries Application Programming Interfaces, along with reading their\ndocumentation, in order to locate the suitable mappings between replacing and\nreplaced methods. In this paper, we introduce RAPIM, a novel machine learning\napproach that recommends mappings between methods from two different libraries.\nOur model learns from previous migrations, manually performed in mined software\nsystems, and extracts a set of features related to the similarity between\nmethod signatures and method textual documentation. We evaluate our model using\n8 popular migrations, collected from 57,447 open-source Java projects. Results\nshow that RAPIM is able to recommend relevant library API mappings with an\naverage accuracy score of 87%. Finally, we provide the community with an API\nrecommendation web service that could be used to support the migration process.\n