SoftLearn: Soft computing for process mining in e-learning

The objective of SoftLearn is the application of soft computing techniques, specifically evolutionary algorithms and fuzzy rules and, also the hybridization of these techniques with those frequently used on process mining, to extract complete, precise, adaptive and hierarchic workflows that model teaching units in e-learning. Moreover, we will propose algorithms for the analysis and evaluation of those workflows and, also, to recommend users for the educational activities.

The SoftLearn project is in line with the orchestrating learning challenge and is intended to contribute in this research topic with tools that let the teachers understand and evaluate the workflow of the activities done in a course. Thus, just using the event logs of the activities undertaken by students through an e-learning platform, SoftLearn will automatically determine the coordination of the activities of the course (flow), who and in which conditions have executed the activities (roles), and what teaching resources have been used in each activity.

Objectives

Automatic generation of the workflow of a teaching unit: we will hybridize the basic process mining techniques with evolutionary algorithms to extract Petri nets that model the teaching units.
Hierarchical organization of the workflows of IMS LD teaching units: starting from flat Petri nets, we will obtain hierarchic Petri nets that formally model the workflow of educational activities as defined in IMS LD.
Analysis and evaluation of hierarchic workflows: we will use classification algorithms based on soft computing (evolutionary algorithms and fuzzy logic) to analyze and evaluate the hierarchic workflows.
Recommendation of relevant users for educational activities execution: we will design recommender algorithms based on hybridization of soft computing and collective intelligence techniques.

<p> The objective of SoftLearn is the application of soft computing techniques, specifically evolutionary algorithms and fuzzy rules and, also the hybridization of these techniques with those frequently used on process mining, to extract complete, precise, adaptive and hierarchic workflows that model teaching units in e-learning. Moreover, we will propose algorithms for the analysis and evaluation of those workflows and, also, to recommend users for the educational activities.</p> The SoftLearn project is in line with the orchestrating learning challenge and is intended to contribute in this research topic with tools that let the teachers understand and evaluate the workflow of the activities done in a course. Thus, just using the event logs of the activities undertaken by students through an e-learning platform, SoftLearn will automatically determine the coordination of the activities of the course (flow), who and in which conditions have executed the activities (roles), and what teaching resources have been used in each activity.</p><ul> <li>Automatic generation of the workflow of a teaching unit: we will hybridize the basic process mining techniques with evolutionary algorithms to extract Petri nets that model the teaching units. </ li> <li>Hierarchical organization of the workflows of IMS LD teaching units: starting from flat Petri nets, we will obtain hierarchic Petri nets that formally model the workflow of educational activities as defined in IMS LD. </ li><li> Analysis and evaluation of hierarchic workflows: we will use classification algorithms based on soft computing (evolutionary algorithms and fuzzy logic) to analyze and evaluate the hierarchic workflows. </ li> <li>Recommendation of relevant users for educational activities execution: we will design recommender algorithms based on hybridization of soft computing and collective intelligence techniques. </li></ul> - Manuel Mucientes Molina - Juan Carlos Vidal Aguiar, Manuel Fernández Delgado, Manuel Lama Penín