In this paper, we consider a minimalistic and behavioristic view of AIS to enable a standardizable mapping of both the behavior of the system and of the learner. In this model, the "learners" interact with the learning "resources" in a given learning "environment" following preset steps of learning "processes." From this foundation, we make several subsequent arguments. (1) All intelligent digital resources such as intelligent tutoring systems (ITS) need to be well-documented with standardized metadata scheme. We propose a learning science extension of IEEE learning object metadata (LOM). specifically, we need to consider cognitive learning principles that have been used in creating the intelligent digital resources. (2) We need to consider AIS as whole when we record system behavior. Specifically, we need to record all four components delineated above (the learners, the resources, the environments, and the processes). We point to selected learning principles from the literature as examples for implementation of this approach. We concretize this approach using AutoTutor, a conversation-based ITS, serving as a typical intelligent digital resource. [This paper was published in: "HCII 2019, LNCS 11597," edited by R. A. Sottilare and J. Schwarz, Springer Nature Switzerland AG, 2019, pp. 204-216.]