Hybrid human-AI driven open personalized education.

Autor/in	Tavakoli, Mohammadreza
Titel	Hybrid human-AI driven open personalized education.
Quelle	Hannover: Gottfried Wilhelm Leibniz Universität (2023), xii, 129 S. PDF als Volltext kostenfreie Datei (1); PDF als Volltext (2); PDF als Volltext kostenfreie Datei (3); PDF als Volltext kostenfreie Datei (4) Link als defekt meldenVerfügbarkeit Dissertation, Gottfried Wilhelm Leibniz Universität Hannover, 2023.
Sprache	englisch
Dokumenttyp	online; Monographie
DOI	10.15488/14084
URN	urn:nbn:de:101:1-2023072002075009351657
Schlagwörter	Bildung; Künstliche Intelligenz; Offene Erziehung; Lernen; Computerunterstützter Unterricht; Dissertation; E-Learning; Personalisierung
Abstract	Attaining those skills that match labor market demand is getting increasingly complicated as prerequisite knowledge, skills, and abilities are evolving dynamically through an uncontrollable and seemingly unpredictable process. Furthermore, people's interests in gaining knowledge pertaining to their personal life (e.g., hobbies and life-hacks) are also increasing dramatically in recent decades. In this situation, anticipating and addressing the learning needs are fundamental challenges to twenty-first century education. The need for such technologies has escalated due to the COVID-19 pandemic, where online education became a key player in all types of training programs. The burgeoning availability of data, not only on the demand side but also on the supply side (in the form of open/free educational resources) coupled with smart technologies, may provide a fertile ground for addressing this challenge. Therefore, this thesis aims to contribute to the literature about the utilization of (open and free-online) educational resources toward goal-driven personalized informal learning, by developing a novel Human-AI based system, called eDoer. In this thesis, we discuss all the new knowledge that was created in order to complete the system development, which includes 1) prototype development and qualitative user validation, 2) decomposing the preliminary requirements into meaningful components, 3) implementation and validation of each component, and 4) a final requirement analysis followed by combining the implemented components in order develop and validate the planned system (eDoer). All in all, our proposed system 1) derives the skill requirements for a wide range of occupations (as skills and jobs are typical goals in informal learning) through an analysis of online job vacancy announcements, 2) decomposes skills into learning topics, 3) collects a variety of open/free online educational resources that address those topics, 4) checks the quality of those resources and topic relevance using our developed intelligent prediction models, 5) helps learners to set their learning goals, 6) recommends personalized learning pathways and learning content based on individual learning goals, and 7) provides assessment services for learners to monitor their progress towards their desired learning objectives. Accordingly, we created a learning dashboard focusing on three Data Science related jobs and conducted an initial validation of eDoer through a randomized experiment. Controlling for the effects of prior knowledge as assessed by the pretest, the randomized experiment provided tentative support for the hypothesis that learners who engaged with personal eDoer recommendations attain higher scores on the posttest than those who did not. The hypothesis that learners who received personalized content in terms of format, length, level of detail, and content type, would achieve higher scores than those receiving non-personalized content was not supported as a statistically significant result. (übernommen).
Erfasst von	Deutsche Nationalbibliothek, Frankfurt am Main
Update	2024/1