Scientometric analysis of emerging trends and research landscape of ERNIE Bot's potentials as an educational tool: A mixed method study of a large language model
This paper presents findings from a mixed-method analysis of a generative AI tool, ERNIE Bot's research status and role in education based on research from 2019–May 2025. Leveraging scientometric analysis, the study explored a combination of bibliometric and discourse analyses to uncover trends and language use. Using the PRISMA model, 147 publications from Scopus database were analyzed and insights from CiteSpace and VOSviewer were combined with thematic analysis. Findings indicate that Chinese researchers, and China-based funding bodies are the most prominent in ERNIE-bot-related research with all the top 5 countries concentrated in the global north. Top themes include ‘Natural Language Model’, ‘Semantics’, and ‘Human’. Topmost publishers were Applied Sciences (Switzerland) , IEEE Access and PLOS ONE with top average citations by Sensors (12.67 citations/publication). The study confirms Bradford and Price laws. ERNIE Bot's application in education has attracted notable attention, with 22.4 % of the analyzed studies focusing on this area, highlighting its strong potential in personalized learning and human-computer interaction. Findings from discourse analysis shows that it’s Chinese origin is often framed as a symbol of contextual leadership and basis for positioning ERNIE Bot as a key player in China's technological competition with global models like ChatGPT. ERNIE Bot's strengths in Chinese language processing and innovations are also highlighted regarding power relations and knowledge hierarchies. However, ERNIE Bot lags global counterparts in overall competitiveness; limitations in data diversity, political censorship, and accessibility further restricts its global applicability. Future research should explore ERNIE Bot's knowledge base, its educational potentials, especially in adaptive learning systems and the significance of contextual leadership as an educational technology.
Authors
- Yan Yang (ORCID: https://orcid.org/0000-0001-5798-3632)
- Bosede Iyiade Edwards (ORCID: https://orcid.org/0000-0002-6922-311X)
Institutions
- Universiti Sains Malaysia (MY)
Publication Details
- Journal
- Social Sciences & Humanities Open
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1016/j.ssaho.2025.101729
- Citations
- 1
- Primary Topic
- AI in Service Interactions
- Type
- article
- Field-Weighted Citation Impact
- 2.48