DIGITAL LIBRARY
LEARNING COMPUTATIONAL THINKING IN A HETEROGLOSSIC ENVIRONMENT: A MULTIMODAL MULTI-AGENT TRANSLANGUAGING APPROACH TO COMPUTATIONAL LITERACIES ACQUISITION VIA GENERATIVE AI
University of Nebraska Lincoln (UNITED STATES)
About this paper:
Appears in: ICERI2024 Proceedings
Publication year: 2024
Pages: 9831-9838
ISBN: 978-84-09-63010-3
ISSN: 2340-1095
doi: 10.21125/iceri.2024.2480
Conference name: 17th annual International Conference of Education, Research and Innovation
Dates: 11-13 November, 2024
Location: Seville, Spain
Abstract:
Although an estimated 43% of the world’s population interact and can build their understanding via two or more languages, learning and schooling often happen exclusively in one language. This renders the task of learning complex concepts more difficult because learners and teachers cannot utilize their full linguistic repertoires during the learning process. Thus, these linguistic restrictions bring about educational inefficiencies across the curriculum from the humanities to computer science. In this presentation and paper we focus specifically on the potential of new technological advances to enable students to learn and express their understanding of computational thinking in multiple languages within the same interface. Across the globe, teachers and students are expected to deepen and broaden their understanding of CT, how it relates to coding, learning programming, and problem solving in general. Herein we describe a heteroglossic learning environment we are tentatively calling Trans-MAMM-CT. This first-iteration design-based research application provides a platform for teachers and students to learn and explore CT using their entire linguistic repertoire in a culturally relevant, personalized, and multimodal way. We assign agents to carry out specific curricular and pedagogical goals and meet linguistic needs in culturally responsive ways. Based on user feedback, agents adjust the nature, modality, and language of their interactions with users. Our presentation and paper respond to a design opportunity during a technological moment when robust real-time multimodal, multi-agent, multilingual systems for student learning and teacher professional development are becoming feasible. Recent advances in the coherence and language capacities of Large Language Models (LLMs) have made it possible to explore how learners might be supported via an environment that affords the use of a range of linguistic and modal avenues. Via design research, we offer a vision, instantiated in the design of Trans-MAMM-CT, of what learning experiences could look like in the 2020s for hundreds of millions of multilingual students.
Keywords:
Computational Thinking, Design-Based Research, Multi-Agent Systems, Multimodal, Translanguaging, Multilingual, Computational Literacies, Learning, Teaching.