Machine Learning based Recommendations to aid Educational Planning and Academic Advising through the Virtual Academic Advisor System.

Inhalt

Literaturnachweis - Detailanzeige

Trefferliste

Autor/Urheber	Vijjapu, Sri Venkata Bhavani Likitha
Titel	Machine Learning based Recommendations to aid Educational Planning and Academic Advising through the Virtual Academic Advisor System.
Quelle	(2019) PDF als Volltext kostenfreie Datei
Sprache	englisch
Dokumenttyp	online; Monographie
Schlagwörter	academic advising; collaborative filtering; educational planning; recommendation engine; sequence classification; synthetic data generation; Computer science; Information science; Educational technology; Computing and software systems
Abstract	Thesis (Master's)--University of Washington, 2019 ; The process of educational planning and academic advising are critical for supporting student retention rate and on-time graduation. Faculty-based advising has been widely adopted by several community colleges because it results in a better advising experience for a student. However, due to lack of resources, outdated technologies, and growing diverse student population, the process increases the workload on already overwhelmed faculty. The role of technology in different areas of education sector is being slowly adopted, for instance, to provide various services such as online classes, registration services, and classwork maintenance. Despite these advancements, it is not being used for the purpose of academic advising. In this thesis, we discuss how machine learning (ML) and other technologies can be used to assist with faculty-based academic advising in higher education planning. The Virtual Academic Advisor (VAA) is an initial software solution to address this problem. Various ML-based techniques such as supervised learning, natural text processing, collaborative filtering, and sequence classification are explored to provide new functionality for the VAA. Collaborative filtering is used to give a recommendation of study plans based on a set of input parameters. Sequence Classification is used to predict possible suitable college majors, based on the course work designated for a student. In addition, we propose a strategy to generate synthetic data, necessary because it is nearly impossible to collect copious amount of real observations necessary for properly training ML-based modules. These various approaches will be integrated into the VAA system, to help faculty with the advising process, saving time for more meaningful conversations with students, and providing students with the ability to explore different educational paths.
Erfasst von	BASE - Bielefeld Academic Search Engine