A Software Ecosystem platform for the development of Recommender Systems

Abstract

Many application domains need to recommend resources for users. The development of solutions focused on the reuse of Recommender System components creates an interesting scenario from a Software Ecosystem perspective. Besides the interaction between actors and technology, a Software Ecosystem for the development of Recommender Systems should allow for integration with systems and platforms that support other ecosystems. The problem addressed by this study is the integration of methods, techniques, and approaches of existing Recommender Systems in a systematic way, facilitating the implementation of new solutions, their reuse and sharing, and the collaboration among the actors involved. This study, therefore, aims to propose R.ECOS, a Software Ecosystem platform to support the development and management of Recommender Systems, allowing for integration between multiple applications and other Software Ecosystems. The evaluation was conducted in two stages. First, two feasibility studies were carried out to validate both technology and architecture. Later, two case studies were conducted in real-world contexts. The results point to the viability of the proposal.