CSEDU_DC 2023 Abstracts


Full Papers
Paper Nr: 5
Title:

Students’ Digital Literacy in the Use of LMS: An Empirical Study Using Gaze-Tracking Technology

Authors:

Tuija Alasalmi

Abstract: In my doctoral research, I examine the manifestation of teacher students’ digital literacy in the use of online learning environments by means of a background survey, self-assessment, gaze-tracking based data and think aloud data/ interviews. Studies of digital literacy are usually based on simple quantitative or qualitative methodology such as users’ self-assessments, standardized tests, observation of activities or the use of protocol analysis, but a mixed method approach using a novel technology provides more reliable information about the phenomenon. The focused topic of the research is the technical aspects of digital literacy, i.e. operational skills and the ability to understand the structure of the online environment, as well as the cognitive strategies that guide students' activities and concrete progress in the online learning management system. Furthermore, it is investigated how the students’ prior experience on ICT and their self-efficacy beliefs influence their mental models when interacting with technology. I analyze the previous research of digital literacy through three main paradigms: the skills perspective, the cognitive approach focusing on interaction and the social perspective emphasizing the role of culture, context and human agency in the use of ICT. Rooted in the tradition of social semiotics theory, the aim of the study is to argue that the individual’s technical, procedural and cognitive competences in digital interaction are not only device-dependent and situation-specific, but also contextually and socially structured.

Paper Nr: 6
Title:

Towards the System-Independent Provision of Personalized Assistance in Learning Management Systems

Authors:

Sebastian Kucharski

Abstract: Intelligent Tutoring Systems (ITS) for Learning Management Systems (LMS) can combine the ITS-specific advantage of user-centered learning with the LMS-specific advantage of providing easy-to-use learning content for numerous users in a single location. These systems typically have two drawbacks. First, they can not be reused for multiple systems. Therefore, many LMS do not contain an ITS and can not adapt the learning content to the user's needs. Second, they lack in an ITS-controlled, didactically utilisable representation of the learning content structure (i.e., they are not structure-aware). Thus, a user can get lost during assistance provision regarding the learning content structure. Therefore, the goal is the investigation of frame conditions for the LMS-independent provision of personalised, structure-aware learning assistance, which is especially challenging due to the heterogeneous, LMS-specific approaches to structure learning content. For this purpose investigation areas and the corresponding research questions towards an LMS-independent, structure-aware ITS are demarcated. This system is intended to overcome the mentioned drawbacks and to be able to take the learning data from different LMS during assistance provision into account. The approach shall be evaluated in terms of applicability and effectiveness considering different state-of-the-art LMS.

Paper Nr: 7
Title:

Supporting Feedback in the HyFlex Model Through a Web-Based Platform

Authors:

Giacomo Cassano

Abstract: The shift towards digitalization in education has led to the development of new models of teaching as well as to a renaissance of existing but not so well known approaches, such as the HyFlex model, which was first proposed by Dr. Brian Beatty in 2006. However, the implementation of the HyFlex model raises several questions related to student interaction, feedback, and learning. Student interaction with the professor during a lecture is crucial for effective learning, as it allows students to ask questions, provide feedback, and gain a deeper understanding of the material. Models like the HyFlex model require tools and support in order to be successful in providing students with a flexible learning experience where interaction is not lost. My research aims to address these challenges by developing a web-based platform, called Evoli, that supports the HyFlex model, improving students’ feedback in a flexible and hybrid environment. In this paper I present my research which tries to answer the research question: "How can a web-based platform support the student-to-teacher feedback process in a HyFlex setting?"

Paper Nr: 8
Title:

A Multi-Modal Chatbot for Education

Authors:

Markos Dimitsas

Abstract: This thesis explores way to effectively segment video lectures and other knowledge assents into coherent topics, and the hypothesis that whether such a knowledge base could support an educational chatbot. Being a part of an ongoing research project I am developing a chatbot that aids the learning process for students of AI, utilizing the plethora of recorded video lectures during the COVID-19 pandemic and other knowledge assets. My main focus however is the application of novel topic segmentation and classification techniques for these assets and to create a corpus of topics in various modalities (video, audio, text), which is a sub-task of the overall project. The chatbot then identifies the most relevant topics, given a student's question, and replies with an automatically composed multi-modal document comprising of relevant video segments, lecture notes, slides and Wikipedia passages. To obtain early gold data on the conversational side, a Wizard-of-Oz experiment is proposed. The quality of the new topic segmentation and classification methods will be assessed against a new corpus of gold data, which will be developed. While there is a substantial body of prior work on video segmentation and some work on education chatbots, I am not aware of any previous educational chatbots that are powered by multi-modal topical segments.