Student dashboard for a multi-agent approach for academic advising

Abstract

The objective of this research is to demonstrate the performance of a new mechanism to improve the advising of students in a nontraditional college environment, specifically the University of Texas at El Paso (UTEP). Minority serving institutions, commuter campuses and institutions with a high percentage of student transfers are unable to keep a tightly controlled cohort of students progressing through the curriculum. Students usually have varied course loads and different priorities due to family, financial needs or other responsibilities. Therefore, there is a need for an individualized approach to advising. The school's administration faces challenges scheduling courses and allocating diminishing resources to satisfy student demand. In addition, faculty needs to assess the efficacy of their curriculum in a program, and collecting longitudinal student data is difficult. A web application system (mobile compatible) using a multi-agent approach has been developed to allow the students (agents) to take more control over their individualized advising. In this context, the student tool becomes an agent, and the school provides the environment with a desirable behavior for the system. This research will identify the school's administrators as the academic control objective and will be referred to as the "Operator". This paper focuses on the agent system by building a dashboard tool that collects each individual student's information regarding their progress through the curriculum in a program, and then generates advising recommendations. The agent logic employs principles used in project management tools designed for resource of schedule optimization. The tool helps students optimize their resources to complete their degree sooner. It provides a visualization map of course sequences, customized for each student based on their history of courses completed and then making advising adjustments that will optimize the time to obtain the degree under a constrained set of resources. At the same time, the agent system provides real-time feedback to the department administration. The second tool is the department administration dashboard that consolidates the collected data from the students through several semesters (historical data) plus the predicted effects of the recommended plans. This enables a better resource allocation from the administration and deeper analysis of the curriculum effectiveness. Previous work has presented some limited insight into the multi-agent approach and the critical path methods. However, the proliferation of mobile devices and Cloud computing enables a larger scale application of the proposed methodology. The results acquired at this point show a very high acceptance of the system by the students. The complete dataset will be discussed extensively in the results section.