This is the parent ticket for the work on the Learning Analytics Processor 1.0. All the work will happen in linked task and feature tickets.
Wiki page with more details:
Complete OAAI Automation and package it as a web service API.
Make OAAI a sustainable open source predictive modeling framework.
Establish a framework for automation and execution of learning analytics models (which is possible to extend with additional model pipelines)
Establish input and output specifications for data used for model processing
API should satisfy the following criteria:
Accept 4 required OAAI data files from the input directory
Validate the data inputs for compliance with OAAI data specifications
Perform data processing and generate predictions using a PMML based model developed by Marist College
Output the results in configurable formats (JSON / CSV / XML)
Provide seamless integration with Sakai CLE