From Colonial Legacies to Linguistic Inclusion: A BERTopic Enhanced Bibliometric Insight Into Global South Higher Education

Abstract

The present study is a comprehensive three-tiered empirical framework that investigates the processes of linguistic diversity and decolonization within the higher education systems of the Global South. It provides a novel complementary application of ARM for Topic Recommender Systems and explores the evolving landscape of scholarly discourse on language justice and educational reform by mapping the findings from Bibliometric Co-occurrence Analysis with LDA and BERTopic modelling respectively. Bibliometric Co-occurrence Analysis was employed to produce seminal relationships between key terms from the extant scholarly literature that subsequently leads to a detailed thematic review. Successive applications of LDA and BERTopic modelling generated a comprehensive thematic analysis that detailed the trends in academic discourse on inclusivity, multilingualism, translanguaging and decolonization of higher education curriculum. Furthermore, by the innovative use of ARM based on author keywords, this study revealed more intricate associations and patterns among research topics, giving important insights into future avenues of research. The analysis was based on data entirely retrieved from the Scopus database. It shows the consensus around the need for reform of language policies in higher education, as well as the transformative potential for advanced computational techniques in analyzing academic narratives. The findings from our methodological exploration, demonstrate the necessity of policy reform in higher education concerning linguistic pluralism and dismantling colonial legacies. This study extends the conversation around the complexities of decolonisation, linguistic equity and justice in higher education in the Global South and beyond.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.