Abstract
[Context] With the increasing industrial demands for seamless exchange of data and services among information systems, architectural solutions are a promising research direction which supports high levels of interoperability at early development stages. [Objectives] This research aims at identifying the architectural problems and before-release solutions of interoperability on its different levels in information systems, and exploring the interoperability metrics and research methods used to evaluate identified solutions. [Methods] We performed a scoping study in five digital libraries and descriptively analyzed the results of the selected studies. [Results] From the 22 studies included, we extracted a number of architectural interoperability problems on the technical, syntactical, semantic, and pragmatic levels. Many problems are caused by systems' heterogeneity on data representation, meaning or context. The identified solutions include standards, ontologies, wrappers, or mediators. Evaluation methods to validate solutions mostly included toy examples rather than empirical studies. [Conclusions] Progress has been made in the software architecture research area to solve interoperability problems. Nevertheless, more researches need to be spent on solutions for the higher levels of interoperability accompanied with proper empirical evaluation for their effectiveness and usefulness.