ABSTRACT
E-Learning, Blended Learning, Massive Online Courses, Distributed Learning, Webinars Hybrid Events and video conferencing are topics treated for decades. Technology-wise, many opportunities were taken and chances were used. However, in many situations, work places, universities and institutions, lectures, courses, trainings, workshops or informative events were still primarily held on-site. All of them, with their specific advantages and disadvantages (e.g., personal contact, networking vs.Time loss, ecological footprint, etc.). First with the Corona Pandemic and the subsequent measures, all foremost the lockdown situations worldwide, remote working, learning and video conferencing experienced a dramatic boost. The option of preferring on-site, personal meetings vanished completely. Due to this profound change in collaborative that had to become common overnight, people were forced to face and adapt to available technologies. The latter, on the other side, also evolved quickly to provide more and more features to improve collaboration, usability and privacy. Formats for lectures, courses, workshops and collaboration also had to be adapted quickly. The problem was that all stakeholders, both lecturers as well as learners could not that easily change their habits, ways of teaching and learning, interacting, and-not to forget-Their didactic concept, methods and materials. A 1:1 transformation was impossible since different technical background knowledge as well as very different technical and spatial conditions lead to great imbalance in quality of teaching, interaction, use of technology and practicability. In order to address these uncertainties, this work provides a basic set of suggestions for lecturers in form of a pattern collection for setting up educational online courses regarding several aspects like common sense, technology and rules of the game. Since patterns are technology-Agnostic and formulated for a broad audience with different professional backgrounds, they qualify as universal format and at the same time keep the validity of time. Even after almost three years with COVID-19, there is still struggle with technology adaption and format generation. The formulated patterns are based on interviews with experts that worked as lecturers, a larger-scale survey including the results from 63 online questionnaires and one focus group. The resulting set of 19 guidelines was then reformulated as patterns in a mid-high maturity state with the perspective of being further supported by new findings and practical experience. The pattern collection is linked to a process model for evolving pattern libraries from previous work. © 2022 Owner/Author.
ABSTRACT
Low-rank tensor factorization or completion is well-studied and applied in various online settings, such as online tensor factorization (where the temporal mode grows) and online tensor completion (where incomplete slices arrive gradually). However, in many real-world settings, tensors may have more complex evolving patterns: (i) one or more modes can grow;(ii) missing entries may be filled;(iii) existing tensor elements can change. Existing methods cannot support such complex scenarios. To fill the gap, this paper proposes a Generalized Online Canonical Polyadic (CP) Tensor factorization and completion framework (named GOCPT) for this general setting, where we maintain the CP structure of such dynamic tensors during the evolution. We show that existing online tensor factorization and completion setups can be unified under the GOCPT framework. Furthermore, we propose a variant, named GOCPTE, to deal with cases where historical tensor elements are unavailable (e.g., privacy protection), which achieves similar fitness as GOCPT but with much less computational cost. Experimental results demonstrate that our GOCPT can improve fitness by up to 2.8% on the JHU Covid data and 9.2% on a proprietary patient claim dataset over baselines. Our variant GOCPTE shows up to 1.2% and 5.5% fitness improvement on two datasets with about 20% speedup compared to the best model. © 2022 International Joint Conferences on Artificial Intelligence. All rights reserved.