Atomic and Molecular Intelligent Tutoring Systems - A New Architecture for Interoperable Tutors as Open Educational Resources

Due to interoperability issues, Intelligent Tutoring Systems are hard to deploy in the different educational platforms available nowadays. This represents a considerable limitation, since tutoring systems require substantial time and a lot of resources to be implemented. This paper describes a new architecture for developing open source and interoperable intelligent tutors. In our approach, "atomic" tutoring systems are grouped to create "molecular" tree structures covering the curriculum of courses. Based on standards, our approach provides interoperable and open source ITSs that can be retrieved from repositories of educational content, grouped when necessary, loaded into different educational platforms and have their code reused for incremental development.