The library is available in three flavors: basic API interaction, logging functionality and the full metacog client library that includes the Playback Mechanism, recording of Training Sessions and Scoring functionality:
|Basic API interaction||yes||yes||yes|
|Logging events to metacog platform||no||yes||yes|
|Recording and playing back of Training Sessions||no||no||yes|
|Visual Scoring of Training Sessions||no||no||yes|
Current version of the library is 3.0.1.
Basic Learning Object instrumentation
In the most basic Use Case, it can be used to make a Learning Object capable of sending learning events to the platform. A Learning Object who had been integrated with the logger in this way is known as an instrumented learning object.
Check the Instrumentation Guide to learn more about the process of integrating the logger within an existing learning object.
Playback-compatible Learning Objects
The Logger Library also includes advanced functionality to make a Learning Object capable of not only sending events to the platform, but also playing them back, and instruct the platform to use a determined set of events as sample data for training Learning Machine Algorithms.
Unlike the Instrumentation use case, that tries to avoid the modification of the original source code of the Instrumented Learning Object as far as possible, this case requires a deep integration of additional metacog components and the modification of the flow of events in the original source code.
Check the Playback Toolbar Guide to learn more about the process of creating learning objects that support playing-back of events and other advanced features.
Scoring Training Sessions
Training Sessions recorded through the Playback Toolbar can also be scored by applying some Rubric's indicators to the event streams. Scored Training Sessions may be used for training Machine Learning Algorithms.
Scoring of Training Sessions requires the existence of at least one Rubric object. Check the rubrics' API endpoint for information about the creation of Rubrics.