Paragraph 1:
The Global South has long faced systemic barriers in the scientific and academic ecosystem, ranging from limited research funding to underrepresentation in high-impact publications. Traditional metrics for assessing research value—such as journal impact factors, citation counts, and institutional rankings—often reflect historical and structural biases favoring institutions in the Global North. These indicators, while widely used, may fail to capture the societal and innovative contributions of research emerging from under-resourced regions. Consequently, there is an urgent need for more inclusive, nuanced, and scalable approaches to evaluating scholarly contributions that recognize both academic excellence and local impact.
Paragraph 2:
Recent advances in artificial intelligence, particularly large language models like ChatGPT, provide a unique opportunity to address these inequities. ChatGPT’s capabilities in natural language understanding, knowledge synthesis, and cross-domain reasoning enable it to analyze and contextualize research outputs beyond traditional quantitative metrics. By leveraging AI-driven assessment tools, stakeholders in the Global South can gain insights into research value, societal relevance, and potential collaborations that were previously difficult to quantify. This paper explores how ChatGPT can be strategically employed to create an equitable research evaluation framework, highlighting methodological innovations, practical applications, and long-term implications for strengthening academic visibility and impact in the Global South.
The evaluation of scientific research has traditionally relied on quantitative metrics such as journal impact factors, citation counts, and institutional rankings. These indicators, while offering standardized benchmarks, often privilege institutions in the Global North due to historical resource imbalances, language dominance, and concentrated publication networks. Researchers in the Global South frequently encounter systemic disadvantages, including limited access to funding, fewer high-impact journal submissions, and underrepresentation in international collaborations. Consequently, conventional metrics may not accurately reflect the true scientific contribution, societal impact, or innovative potential of research produced in these regions.
The Global South, encompassing regions in Africa, Latin America, the Middle East, and parts of Asia, exhibits substantial heterogeneity in research capacity, infrastructure, and funding. Despite these disparities, local researchers frequently address pressing societal challenges such as public health crises, climate adaptation, and sustainable development. Their work often provides practical solutions with immediate local relevance, yet remains undervalued in mainstream evaluation systems that prioritize citations, journal prestige, and global visibility. For instance, studies in local languages, community-driven projects, or applied research may be systematically overlooked, despite their high societal impact. This mismatch underscores the need for more context-sensitive and multidimensional evaluation frameworks that recognize both scientific rigor and social relevance.
Recent scholarship has highlighted the limitations of traditional bibliometric indicators. The reliance on citation counts can create a feedback loop, amplifying the visibility of already well-resourced institutions while marginalizing emerging research communities. Similarly, journal impact factors are often criticized for reflecting editorial policies rather than the intrinsic quality of individual studies. These limitations have prompted scholars to explore alternative evaluation strategies, including altmetrics, expert peer assessments, and multidimensional impact evaluations. However, these methods can be labor-intensive, inconsistent across disciplines, and still constrained by existing power structures in global academia.
Artificial intelligence (AI), particularly large language models (LLMs) like ChatGPT, has emerged as a promising tool to complement and enhance traditional research evaluation. LLMs are capable of analyzing vast volumes of academic text, synthesizing knowledge across disciplines, and providing semantic insights that go beyond simple citation metrics. For example, AI-driven analyses can evaluate novelty, identify research trends, and estimate societal relevance by contextualizing publications within broader global and local knowledge networks. Moreover, these models can process multilingual corpora, enabling equitable evaluation of research in non-English languages—a critical feature for the Global South, where locally relevant scholarship often remains underrepresented in English-dominated databases.
Early experiments demonstrate that AI models can assist in peer review, summarize research impact, and highlight interdisciplinary connections that traditional bibliometrics may overlook. By integrating AI-driven assessment with conventional metrics, institutions and policymakers can develop more nuanced and scalable evaluation frameworks. This hybrid approach not only enhances transparency and efficiency but also mitigates biases inherent in conventional evaluation systems. Importantly, the deployment of AI in research evaluation requires careful consideration of model biases, data coverage, and interpretability, ensuring that the Global South benefits from rather than is disadvantaged by these technological tools.
In summary, the scientific landscape of the Global South is characterized by both significant potential and structural challenges. Traditional evaluation metrics, though widely used, are insufficient to capture the multidimensional value of research in these regions. AI, exemplified by ChatGPT, offers an innovative pathway to enhance assessment practices, providing richer, context-aware insights that can empower underrepresented researchers and institutions. This section establishes the foundation for exploring methodological frameworks and strategic applications of ChatGPT for research evaluation in the Global South, setting the stage for subsequent sections on methods, strategy, and practical implementation.
The application of ChatGPT to research evaluation represents a paradigm shift in assessing scholarly contributions, particularly for underrepresented regions in the Global South. Traditional metrics, while providing numerical indicators of impact, often fail to capture the nuanced dimensions of research quality, interdisciplinary significance, and societal relevance. ChatGPT, as a large language model, offers advanced capabilities in natural language understanding, semantic analysis, and knowledge synthesis, enabling a richer, multidimensional evaluation framework. This section outlines a comprehensive methodological approach to leveraging ChatGPT in research assessment, emphasizing fairness, transparency, and interpretability.
At the core of this framework is the conceptual shift from purely quantitative evaluation to hybrid, context-aware assessment. ChatGPT can analyze textual content across academic papers, reports, and policy documents to extract key indicators of research value. These include:
Novelty and originality: Identifying unique contributions within a research domain, including innovative methodologies or conceptual frameworks.
Interdisciplinary relevance: Mapping connections between research outputs and other scientific domains, highlighting cross-disciplinary influence.
Societal impact: Evaluating potential applications in addressing local and global challenges, particularly those relevant to the Global South.
Language and accessibility: Assessing the clarity, inclusivity, and accessibility of research outputs, especially in non-English languages.
By integrating these qualitative dimensions with traditional metrics (citations, journal impact), ChatGPT enables a more holistic assessment of research value that is both rigorous and context-sensitive.
Effective AI-driven evaluation requires comprehensive and high-quality data. The first step involves compiling a corpus of research outputs, which may include:
Peer-reviewed journal articles
Conference proceedings
Policy briefs and technical reports
Preprints and institutional working papers
Data preprocessing is critical to ensure consistency and reliability. Key steps include:
Text normalization: Removing formatting inconsistencies, special characters, and redundant metadata.
Language detection and translation: For multilingual corpora, non-English texts are translated or processed with multilingual models to ensure inclusivity.
Metadata enrichment: Incorporating contextual information such as author affiliations, funding sources, and research domains.
This preprocessing ensures that ChatGPT can accurately parse and interpret the research content, minimizing biases introduced by inconsistent or incomplete data.
Once the dataset is prepared, ChatGPT can be applied through a series of analytical procedures:
Content summarization: Generating concise summaries of research articles to identify key contributions, methodological approaches, and findings.
Semantic similarity analysis: Comparing new research outputs with existing literature to assess novelty and originality.
Citation context analysis: Examining the language and context in which research is cited to evaluate influence and interdisciplinary relevance.
Impact prediction: Using prompt engineering to estimate potential societal impact, particularly in local or regional contexts relevant to the Global South.
These procedures allow evaluators to move beyond raw citation counts and journal rankings, providing a richer understanding of the research’s significance.
Despite its capabilities, ChatGPT is not free from biases inherent in training data and knowledge representation. Methodological safeguards include:
Diverse data sourcing: Incorporating publications from a wide range of institutions, languages, and disciplines to reduce Northern-centric bias.
Prompt refinement and iterative testing: Using carefully designed prompts and multiple evaluation iterations to enhance consistency and reliability.
Transparency and interpretability: Documenting the evaluation process and providing human-readable explanations for AI-driven assessments.
Such safeguards are particularly important for equitable evaluation in the Global South, ensuring that AI amplifies rather than entrenches existing disparities.
The methodological framework does not seek to replace traditional evaluation entirely but to complement it. ChatGPT-driven insights can be combined with bibliometrics and peer reviews to create a hybrid assessment model. This integrated approach allows decision-makers to:
Identify high-potential research that might be undervalued in traditional metrics
Recognize applied and socially relevant contributions
Support funding allocation, academic promotions, and collaborative initiatives based on multidimensional evaluation
Finally, the framework emphasizes iterative validation to ensure robustness. Evaluators can compare ChatGPT-generated assessments with expert judgments, longitudinal research outcomes, and societal impact indicators. Feedback loops are critical for refining prompts, adjusting analytical procedures, and enhancing model performance over time. This continuous improvement ensures that AI-driven evaluation remains aligned with the evolving needs of the Global South and global research community.
Summary: This methodological framework establishes a structured, transparent, and multidimensional approach to research evaluation using ChatGPT. By combining content analysis, semantic insights, and societal impact assessment, it empowers underrepresented researchers while complementing traditional metrics. Key components—data preprocessing, analytical procedures, fairness safeguards, integration with conventional indicators, and iterative validation—form the foundation for a strategic AI-driven evaluation system tailored to the Global South.
The Global South encompasses a diverse set of countries characterized by varying levels of economic development, research infrastructure, and academic visibility. While these regions are increasingly contributing to global scientific output, systemic inequities continue to limit recognition, funding opportunities, and international collaboration. Leveraging ChatGPT as a strategic tool for research evaluation can help mitigate these disparities and inform policies that strengthen the capacity, visibility, and impact of science in the Global South.
A central challenge for the Global South is the underrepresentation of local research in global knowledge networks. Many high-impact journals and indexing databases are biased toward English-language publications and institutions with established reputations in the Global North. ChatGPT can serve as an intermediary by analyzing and summarizing research outputs in multiple languages, providing structured assessments that highlight both scientific rigor and societal relevance.
By doing so, AI-driven evaluation systems can:
Identify high-quality but underrepresented research: Flagging contributions that may be overlooked by traditional metrics.
Facilitate multilingual accessibility: Translating, summarizing, and contextualizing research to broaden reach and comprehension.
Support equitable funding allocation: Providing policy-makers and funding agencies with AI-generated insights that highlight local impact and innovation potential.
This approach ensures that research from under-resourced institutions receives fair consideration in global evaluation systems, promoting inclusivity and diversity in knowledge production.
Beyond individual research evaluation, ChatGPT can assist policymakers and academic leaders in strategic planning. By synthesizing trends across disciplines, identifying emerging research clusters, and mapping local challenges to global scientific priorities, AI can guide investment decisions in a targeted and efficient manner. For example, governments and research councils can use ChatGPT to:
Determine high-impact research areas aligned with national development goals
Optimize resource allocation for interdisciplinary initiatives
Anticipate future societal needs, such as public health, climate adaptation, or digital infrastructure
This strategic alignment ensures that limited resources are deployed effectively, maximizing both scientific output and social relevance.
ChatGPT can also play a vital role in capacity building within the Global South. By providing real-time assistance in literature review, grant proposal drafting, and policy analysis, AI can enhance the productivity and skill sets of researchers. Additionally, AI-driven platforms can facilitate the creation of collaborative knowledge networks:
Connecting researchers across institutions, countries, and disciplines
Promoting mentorship and skill sharing through AI-curated content
Encouraging regional research collaborations that address locally relevant challenges
Such initiatives not only strengthen individual research capacity but also foster robust, self-sustaining academic ecosystems.
The insights generated by ChatGPT can inform evidence-based policy decisions in multiple domains: research funding, academic promotions, and international collaboration strategies. Policymakers can utilize AI-driven assessments to:
Identify systemic gaps in research infrastructure or training
Evaluate the societal impact of funded projects
Develop incentives that encourage research addressing local and regional challenges
By integrating AI insights into policy frameworks, governments and institutions can implement strategies that enhance both academic excellence and social utility, ensuring that research serves the broader needs of society.
While ChatGPT offers substantial strategic advantages, careful consideration is required to avoid reinforcing existing inequities. Potential risks include algorithmic bias, overreliance on AI-generated outputs, and the exclusion of non-digital or unpublished research. Strategic deployment in the Global South requires:
Transparent methodologies and documentation of AI-driven evaluations
Integration with human oversight to validate assessments
Continuous monitoring and adjustment to ensure alignment with local priorities and ethical standards
This ensures that AI serves as a supportive tool rather than an authoritative gatekeeper, empowering rather than limiting researchers in the Global South.
Summary:
From a strategic perspective, ChatGPT can significantly enhance research evaluation and policy-making in the Global South. By improving visibility, guiding investment, strengthening capacity, and informing evidence-based policies, AI-driven tools create a more equitable and effective academic ecosystem. Ethical deployment, transparency, and human oversight remain essential to ensure that these technologies genuinely support the growth, influence, and societal relevance of research in underrepresented regions.
To illustrate the practical utility of ChatGPT in evaluating research value within the Global South, several case studies highlight how AI-driven methodologies can complement traditional metrics and provide actionable insights. These examples emphasize context-sensitive assessment, interdisciplinary mapping, and societal impact evaluation.
A research consortium focused on infectious diseases in Sub-Saharan Africa faced challenges in gaining international visibility due to publications primarily in French and Portuguese. Using ChatGPT, researchers were able to:
Translate key findings into English while maintaining semantic accuracy
Summarize research contributions, highlighting innovative diagnostic methodologies and local impact
Compare outputs with global research trends to identify novel approaches
The AI-assisted evaluation revealed that several studies, though under-cited in conventional metrics, provided critical solutions to public health challenges. This allowed funding agencies to recognize and support high-impact work that might have been otherwise overlooked.
A Latin American initiative aimed at developing climate-resilient crops produced numerous policy briefs and field studies with high local relevance but low international citation counts. ChatGPT was used to:
Generate concise summaries for decision-makers and policymakers
Assess potential societal impact by analyzing content against regional agricultural needs
Identify cross-disciplinary connections with environmental science, economics, and social policy
By integrating these AI-generated insights with conventional bibliometrics, stakeholders could prioritize research areas with the greatest local relevance and cross-sectoral benefits, thereby maximizing both academic and societal impact.
In Southeast Asia, several small universities collaborated on digital innovation projects, including AI in education and renewable energy technologies. These projects often faced difficulty demonstrating global relevance due to publication in regional journals. ChatGPT facilitated:
Semantic analysis of research papers to highlight originality and interdisciplinary contributions
Identification of potential international collaborators and citation networks
Creation of AI-assisted visual maps showing thematic clusters and emerging trends
The AI-supported approach allowed these universities to showcase the broader relevance of their work, attracting international partners and fostering knowledge exchange.
Across these case studies, several common insights emerge:
Enhanced Recognition of Local Impact: ChatGPT can reveal the societal and innovative contributions of research that conventional metrics often overlook.
Multilingual and Interdisciplinary Integration: AI can bridge language barriers and highlight cross-domain relevance, which is essential for underrepresented regions.
Actionable Insights for Policymakers: Summarization, trend analysis, and impact prediction allow stakeholders to make evidence-based decisions regarding funding, collaboration, and strategic investment.
Complementary to Traditional Metrics: While AI provides rich qualitative insights, its integration with conventional bibliometric indicators ensures a balanced and robust evaluation framework.
These practical applications demonstrate that ChatGPT is not merely a research assistant but a strategic tool capable of reshaping evaluation practices, promoting equity, and amplifying the voices of Global South researchers. By contextualizing research contributions within local and global frameworks, AI-driven evaluation facilitates informed decisions that advance both scientific innovation and societal well-being.
While ChatGPT presents promising opportunities for research evaluation in the Global South, its deployment is accompanied by multiple challenges that must be carefully considered. These challenges span technical constraints, ethical considerations, and institutional limitations, each of which can influence the effectiveness, fairness, and credibility of AI-driven assessment.
Data Quality and Coverage:
AI-driven evaluation relies on large, high-quality datasets. In the Global South, research outputs may be dispersed across local journals, institutional repositories, or unpublished reports. Limited digital accessibility and inconsistent metadata can compromise ChatGPT’s ability to process and analyze these documents accurately. Incomplete or biased datasets risk reinforcing visibility disparities rather than alleviating them.
Language and Semantic Limitations:
Although ChatGPT can process multiple languages, its performance varies across linguistic contexts. Local languages, dialects, and domain-specific terminology may be underrepresented in the training data, leading to inaccuracies in semantic interpretation or summarization. Consequently, research written in under-resourced languages may be undervalued or misinterpreted.
Scalability and Computational Resources:
Deploying ChatGPT at scale requires significant computational resources, including high-performance GPUs and reliable cloud infrastructure. Many institutions in the Global South may lack the technical capacity to implement AI-driven evaluation effectively, creating a gap between potential and actual utilization.
Model Bias and Hallucinations:
ChatGPT may generate outputs influenced by inherent biases in its training data or produce “hallucinations”—plausible but inaccurate statements. These risks are especially critical in research evaluation, where misinterpretation of scientific contributions can have tangible consequences, such as misallocation of funding or misjudgment of academic merit.
Equity and Fairness:
AI models can inadvertently reproduce historical and structural biases present in global academic databases. Without careful design and oversight, AI-driven evaluation may favor institutions and researchers already visible in international networks, undermining the goal of promoting equity in the Global South.
Transparency and Accountability:
The opacity of AI decision-making presents challenges for accountability. Evaluators and policymakers need clear explanations of AI-generated assessments to trust and act upon them. Lack of interpretability could lead to skepticism or resistance among researchers and stakeholders.
Privacy and Data Security:
Evaluating research outputs often requires handling sensitive information, including unpublished manuscripts, grant applications, or institutional reports. Ensuring data privacy, confidentiality, and compliance with ethical standards is critical to maintain trust in AI-driven evaluation systems.
Integration with Existing Systems:
Many academic institutions rely on traditional evaluation processes, including peer review, citation metrics, and tenure committees. Introducing AI-driven tools requires institutional buy-in, training, and workflow adjustments. Resistance from stakeholders accustomed to conventional metrics can slow adoption.
Standardization and Validation:
The absence of universally accepted standards for AI-assisted research evaluation complicates implementation. Validation of ChatGPT-generated assessments against peer evaluations, societal impact measures, and longitudinal outcomes is necessary to establish credibility and reliability.
Resource and Capacity Disparities:
Institutions in the Global South vary widely in infrastructure, funding, and technical expertise. Without equitable access to AI tools and training, disparities may widen, with only well-resourced institutions benefiting from AI-driven evaluation, while smaller or rural institutions remain marginalized.
To address these challenges, several mitigation strategies can be implemented:
Hybrid Evaluation Models: Combining AI-driven insights with human peer review ensures balanced and validated assessments.
Inclusive Training Data: Curating diverse, multilingual datasets representing under-resourced regions reduces bias and enhances model accuracy.
Capacity Building: Investing in training, infrastructure, and technical support enables broader adoption of AI evaluation tools.
Transparency and Documentation: Clear reporting of methods, prompts, and decision logic enhances trust and interpretability.
Ethical Oversight: Institutional review boards or ethics committees can monitor AI-driven assessment processes to ensure compliance with equity, privacy, and accountability standards.
Summary:
While ChatGPT offers transformative potential for research evaluation in the Global South, its adoption is not without risks. Technical limitations, ethical dilemmas, and institutional barriers must be proactively addressed to ensure equitable, accurate, and trustworthy assessment outcomes. Through hybrid evaluation models, inclusive data practices, capacity building, and ethical oversight, AI-driven tools can be harnessed responsibly to complement traditional evaluation methods and promote global research equity.
The integration of ChatGPT into research evaluation presents a transformative opportunity for the Global South, offering both immediate benefits and long-term strategic potential. Looking ahead, several key directions can enhance the effectiveness, inclusivity, and impact of AI-driven evaluation systems.
Future applications of ChatGPT should prioritize localized assessment frameworks that reflect regional priorities, languages, and cultural contexts. By incorporating context-aware prompts, regional databases, and domain-specific corpora, AI models can provide evaluations that are more relevant to local challenges and development goals. This approach ensures that research addressing local societal issues—such as public health, sustainable agriculture, or climate adaptation—is accurately recognized and valued, complementing global academic metrics.
Global South research is often published in diverse languages and spans multiple disciplines. ChatGPT’s multilingual capabilities can be expanded through specialized training on underrepresented languages, enabling comprehensive analysis of local scholarship. Additionally, AI-driven tools can map interdisciplinary connections, revealing hidden synergies between seemingly disparate research domains. This facilitates collaboration, identifies emerging research trends, and highlights novel approaches that traditional bibliometric systems may overlook.
ChatGPT can play a pivotal role in shaping evidence-based policy and funding strategies. Future development could focus on integrating AI-driven insights with institutional planning tools to:
Predict emerging areas of high societal impact
Recommend strategic investments in under-resourced fields or institutions
Monitor research progress and outcomes in real-time
Such applications would enable governments, funding agencies, and academic institutions to make informed decisions that maximize both scientific and societal returns.
AI-driven platforms can facilitate the creation of collaborative research networks within and across regions in the Global South. By connecting researchers based on thematic relevance, methodological complementarity, and societal impact potential, ChatGPT can foster mentorship, knowledge sharing, and joint research initiatives. Future innovations may include AI-mediated matchmaking for research collaborations, virtual think tanks, and adaptive training modules tailored to local needs.
The deployment of ChatGPT in research evaluation should emphasize iterative improvement. Continuous feedback loops, where AI-generated assessments are compared with peer review and longitudinal impact outcomes, will enhance model accuracy and reliability. Over time, these adaptive systems can refine evaluation criteria, identify biases, and evolve to better serve the diverse research ecosystems of the Global South.
Future directions must prioritize ethical AI governance to ensure fairness, transparency, and accountability. Developing standardized guidelines for AI-driven evaluation, monitoring algorithmic bias, and ensuring inclusive access to AI tools are essential steps. By embedding ethical oversight within technological frameworks, the Global South can harness AI responsibly, amplifying underrepresented voices without reinforcing structural inequities.
Looking ahead, the convergence of AI, multilingual capability, and strategic evaluation frameworks could transform research ecosystems in the Global South. ChatGPT has the potential not only to supplement traditional metrics but also to redefine what constitutes “research value,” emphasizing societal relevance, interdisciplinary innovation, and regional impact. This transformation could enhance global research equity, foster local innovation, and strengthen the contribution of the Global South to worldwide scientific progress.
Summary:
The future of AI-driven research evaluation in the Global South lies in localized, multilingual, and interdisciplinary approaches that integrate ChatGPT with evidence-based policymaking, collaborative networks, and continuous ethical oversight. By aligning AI capabilities with regional priorities, these innovations have the potential to reshape research evaluation, amplify underrepresented contributions, and drive sustainable scientific development.
This study has explored the transformative potential of ChatGPT in evaluating research value within the Global South. Traditional metrics, while widely used, often fail to capture the nuanced contributions of underrepresented regions, including societal relevance, interdisciplinary innovation, and local impact. By leveraging the advanced language understanding and knowledge synthesis capabilities of ChatGPT, researchers and policymakers can gain a richer, more context-sensitive perspective on scientific contributions.
The methodological framework presented demonstrates how AI can complement conventional bibliometrics, enabling hybrid evaluation systems that balance quantitative metrics with qualitative insights. Strategic applications of ChatGPT—from multilingual assessment to trend mapping and impact prediction—can enhance visibility, guide resource allocation, and foster collaboration across institutions and disciplines. Case studies illustrate its practical value in identifying high-impact research, supporting policy decisions, and promoting equity in research evaluation.
While technical, ethical, and institutional challenges remain, proactive mitigation strategies—including inclusive datasets, human oversight, and transparent evaluation processes—ensure responsible deployment. Looking forward, continuous refinement, ethical governance, and localized, interdisciplinary frameworks will further amplify the benefits of AI-driven evaluation. In sum, ChatGPT offers a strategic pathway for the Global South to redefine research value, strengthen academic ecosystems, and achieve greater global scientific recognition, bridging long-standing inequities and empowering local innovation.
Hicks, D., Wouters, P., Waltman, L., de Rijcke, S., & Rafols, I. (2015). Bibliometrics: The Leiden Manifesto for research metrics. Nature, 520(7548), 429–431. https://doi.org/10.1038/520429a
Mugwagwa, J., & Edwards, P. (2013). Science, technology and innovation in Africa: Policies for inclusive development. Technological Forecasting and Social Change, 80(5), 791–804. https://doi.org/10.1016/j.techfore.2012.09.005
Bornmann, L. (2014). Do altmetrics point to the broader impact of research? An overview of benefits and disadvantages of altmetrics. Journal of Informetrics, 8(4), 895–903. https://doi.org/10.1016/j.joi.2014.09.005
Waltman, L. (2016). A review of the literature on citation impact indicators. Journal of Informetrics, 10(2), 365–391. https://doi.org/10.1016/j.joi.2016.02.007
Floridi, L., Chiriatti, M., & Taddeo, M. (2023). The ethics of AI in research evaluation: Opportunities and challenges. AI & Society, 38, 1071–1085. https://doi.org/10.1007/s00146-023-01624-3
Sugimoto, C. R., Work, S., Larivière, V., & Haustein, S. (2017). Scholarly use of social media and altmetrics: A review of the literature. Journal of the Association for Information Science and Technology, 68(9), 2037–2062. https://doi.org/10.1002/asi.23833
UNESCO. (2021). Global Education Monitoring Report: Inclusion and equity in education. Paris: UNESCO Publishing. https://unesdoc.unesco.org/ark:/48223/pf0000375709
van Dijk, D., Manor, O., & Carey, L. (2020). AI in research assessment: Current applications and future perspectives. Scientometrics, 125(3), 2211–2236. https://doi.org/10.1007/s11192-020-03506-4
Abebe, R., Gay, G., & Shneiderman, B. (2021). Supporting Global South research with AI-driven tools: Opportunities and constraints. Computers & Education, 170, 104220. https://doi.org/10.1016/j.compedu.2021.10422
ChatGPT. (2023). GPT-4 Technical Report. OpenAI. https://cdn.openai.com/papers/gpt-4.pdf
