Online instruction is facilitated using appropriate software tools that enhance the functionality of the learning platform and broaden the instructional variety that can be materialized for the purpose of learning. It seems though that aspects of the instruction such as the identification of behavioural characteristics of learners, the assessment of effectiveness of learning resources and the awareness for instructional elements and procedures (that closely related to its success) are currently inadequately supported.The network of interactions which evolves when participants perform learning activities can be represented using appropriately defined graph models capable of expressing the aforementioned properties of the instruction. The visual expressiveness of the graph model impacts its effectiveness and greatly relates to the algorithms used for laying out its elements and properties. We describe the architecture of a subsystem which aspires to assist online learning instructors by parsing system log-data to descriptive visual representations using appropriate graph layout methods.