Exploiting Computational Linguistics and Software-Defined Networking to enhance Corpus-Based Curriculum Design: a network analysis approach
Abstract
Knowledge from computational linguistics and digital pedagogy suggests that curricular models lose contextual alignment over dynamic learning environments, but the mechanisms underlying the co-evolution of linguistic structures and instructional frameworks are not completely understood. Drawing on the literature of network analysis, software-defined architectures, and corpus-driven education, this paper constructs and verifies the multi-layered SEM model of language node interaction–semantic clustering–pedagogical adaptation, with curricular modules as the objects. A conceptual mapping framework is presented to elaborate the distinct qualities of communication flows occurring in learner-data, teacher-content, and platform-mediated interactions and to analyze their emergent properties based on network topologies and corpus annotation experiences. The results show that semantic coherence metrics have a significantly positive effect on learner engagement, which in turn significantly promotes the curriculum adaptability, but infrastructure-level parameters do not significantly affect the linguistic diversity. The structural equation modeling analysis reveals that the higher the modularity of interaction nodes, the more significant the effect of content synchronization on learning continuity and the more prominent the influence of dynamic routing on the instructional relevance. Furthermore, it may also support practical guidance regarding when and how various forms of network-informed encounters could be incorporated into competency-based learning pathways. Finally, the authors discussed the implications of the research results for curriculum engineering and education technology deployment.