Purpose This study examined how artificial intelligence (AI) tools can be integrated into knowledge management (KM) systems in higher education institutions and assessed their impact on key KM objectives and learning-related outcomes. Design/methodology/approach In terms of purpose, the research was applied and developmental. Methodologically, a qualitative approach was adopted. The study participants consisted of faculty members from Iranian universities. Sampling was conducted using the snowball technique. A total of 11 faculty members participated in the study. The primary data collection instrument was the semi-structured interview. Using a fuzzy cognitive mapping (FCM) approach, the study modeled the relationships between AI technologies, KM variables, and KM functions. Findings The results indicated that AI tools have a significant positive effect on KM outcomes in higher education. Natural language processing (NLP) showed the strongest influence on personalized learning, while intelligent assistants and chatbots had the greatest impact on real-time feedback, decision-making, and pattern recognition. Overall, AI-driven KM systems enhanced collaboration, learning outcomes, and knowledge sharing, while improving the efficiency of decision-making processes. Originality/value This study contributes to the higher education literature by offering a systemic, AI-driven KM framework based on FCM. It provides practical insights for university leaders and policymakers on utilizing AI to strengthen KM practices, foster innovation, and support collaborative and data-informed decision-making in academic institutions.
Purpose This study assesses the extent to which large language models (LLMs) reproduce the IELTS Writing Task 2 band scores assigned by professional IELTS examiners. It also evaluates whether incorporating the official IELTS band descriptors into the grading prompt improves scoring accuracy and agreement. Design/methodology/approach Six LLM routes (GPT-4o, incept5/llama3.1-claude, Gemma 3, Aya, deepseek-local and deepseek-remote) graded a stratified 100-essay subset from a public IELTS benchmark containing dataset-provided band labels. Each essay was graded 10 times under No Rubrics (NR) and With Rubrics (WR). Scores were treated as single-label multi-class ordinal classifications over the 4.0–9.0 half-band scale and evaluated using accuracy, precision, recall, F1-score, Cohen's kappa, Fleiss' kappa, grading gaps, directional bias and run-to-run stability. Findings Rubric guidance improved average performance and most model-level classification metrics, with cross-model mean gains of 8.8% in accuracy, 14.9% in precision, 23.2% in recall and 18.7% in F1-score. Within the parseable-response subset, GPT-4o achieved the highest reported accuracy and the strongest agreement with the dataset-provided labels; however, its results rest on a substantially smaller set of valid responses than the other routes, and agreement remained moderate at best. Inter-model agreement was negligible; most models inflated scores, and in this model pool, the local/Ollama routes showed larger gaps and instability. The results do not support unsupervised high-stakes marking. Originality/value The study provides an applied, risk-aware evaluation protocol for LLM-assisted writing assessment by jointly reporting correctness, agreement, directional bias and run-to-run stability rather than accuracy alone. It supports rubric-anchored, human-in-the-loop LLM use for formative feedback, triage and moderation, not examiner replacement.
Purpose This study examines the association between graduates' retrospective assessments of an experiential learning curriculum and their self-reported perceived competitiveness, while separately describing employment relevance and monthly income. The study treats the competence-wage paradox as a descriptive and interpretive pattern rather than an empirically tested wage relationship. Design/methodology/approach A quantitative, cross-sectional ex post facto design was employed. Survey data from 55 graduates of an Islamic education management programme were analysed using partial least squares structural equation modelling (PLS-SEM) and descriptive statistics. PLS-SEM was used to assess the association between the experiential learning curriculum and perceived graduate competitiveness, while employment and income were analysed descriptively. Findings Graduates' assessments of the experiential learning curriculum were positively associated with their self-reported perceived competitiveness. Most respondents reported being employed and working in education-related fields, while 63.6% reported monthly income below IDR 1,000,000. The coexistence of perceived competitiveness, job relevance, and comparatively low reported income is described as a context-specific competence-wage paradox. This pattern is interpretive and descriptive; because income was not included in the structural model, the study does not test whether competitiveness affects earnings. Originality/value This study extends graduate employability research by distinguishing perceived competitiveness and employment relevance from reported income in a context-specific cohort of Islamic education management graduates. It also demonstrates the importance of assessing graduate outcomes beyond employment rates by considering job relevance and income.
Purpose This research explores tertiary students' career preparedness in Bangladesh. In addition, this research assesses the extent to which the Japanese University-Industry Linkage model can be implemented to advance their employability within the Bangladeshi higher education context. Design/methodology/approach This research employed a mixed-methods design with deductive reasoning. This research used a structured questionnaire to survey 200 tertiary students from public and private universities in Bangladesh. In addition, 20 Key Informant Interviews were conducted with professionals using a semi-structured questionnaire. Findings The moderate level of career readiness among Bangladeshi students persisted. At the same time, some students have substantially higher expectations for their career prospects. However, they lack important skills such as communication, digital literacy and problem-solving. On the other side of the coin, the institutionalized career centers, compulsory internships, active industry involvement in curriculum development and tight collaboration between universities and industry, supported by long-term policies, have made the Japanese approach integrated and comprehensive. Practical implications At the macro level, it calls for the proliferation of dedicated career consultancy centers and the inclusion of credit-bearing internships in academic programs. Additionally, at the institutional level, industry leaders should support curriculum design to align curricula with market demand. Together, these multi-level strategies establish a structured framework for sustainable career preparation and workforce transition. Originality/value This research, based on empirical data, is the first and foremost to examine the Bangladeshi context through the lens of Japan's UIL model of higher education and to provide cross-cultural policy lessons.
Purpose This study tests whether computational text indicators of epistemic norms add out-of-sample predictive value for normalized academic results beyond a weak baseline of directly observed household/resource indicators. Design/methodology/approach Data comprise 263 students observed in two waves across five sites. Prediction uses 263 follow-up records. The baseline contains seven observed indicators. Archived epistemic counts were generated through pre-assigned mappings for closed items and LDA/topic-model-assisted processing for open responses. A deterministic 10-fold split (seed 60126) evaluated raw counts, within-fold PCA, and fixed outcome-independent ENA coordinates using pooled OOF R2, RMSE and MAE. Findings No epistemic specification reliably improved the baseline. Pooled OOF R2 was −0.009 (M0), −0.065 (M1), −0.032 (M1-PCA) and −0.008 (M2). M1 significantly worsened MAE after Holm correction; H1 and H2 were not supported. Research limitations/implications The baseline is weak, no prior-achievement variable is available, corpus-level reliability of automated epistemic variables is unknown, ENA coordinates were fixed before resampling, and program representation is imbalanced. Practical implications The indicators should not be used for high-stakes decisions without prospective measurement, subgroup-fairness and instructional-utility validation. Originality/value The study provides an auditable reporting object linking feature provenance, leakage risk, resampling controls and reporting requirements, and shows why null or adverse predictive results should be retained.
