Designing Intelligent Learning Ecosystems: The Role of Artificial Intelligence and Blended Learning in Enhancing Digital Education Quality

Figure 1. Proposed intelligent learning ecosystem model

The qualitative findings are summarized in Table 1. The thematic analysis identified 11 subthemes grouped under the three main dimensions. Within the blended learning dimension, the extracted components included flexibility, interaction, personalization, and infrastructure and access. These results indicate that effective blended learning is not limited to combining online and face-to-face delivery modes, but rather depends on the capacity of the learning environment to provide adaptable pathways, active communication, individualized learning opportunities, and sufficient technological support. In this respect, blended learning emerged as the foundational educational mechanism through which digital education quality can be enhanced.

1. Table 1. Summary of qualitative results and final thematic structure

Table note. The file reports that the thematic analysis yielded 96 selected speech excerpts, organized into 11 subthemes and 3 main dimensions.

The second main dimension consisted of artificial intelligence capabilities, including educational data analysis, intelligent recommendation, intelligent support, and automated assessment. These findings suggest that artificial intelligence contributes to digital education not merely through automation, but through more adaptive, responsive, and data-informed learning processes. Educational data analysis enables monitoring of learner behavior and performance patterns; intelligent recommendation supports personalized content delivery; intelligent support facilitates timely guidance and sustained motivation; and automated assessment improves the efficiency and consistency of feedback and evaluation processes. Together, these capabilities form the technological infrastructure of the proposed ecosystem.

The third dimension, e-learning quality, was represented by learner satisfaction, learning effectiveness, and educational interaction. These indicators frame educational quality as a multidimensional and learner-centered construct. The findings therefore show that the proposed model does not treat technological integration as an end in itself; instead, it emphasizes the improvement of meaningful educational outcomes and experiences.

The quantitative findings further confirmed the validity of the proposed model. As reported in Table 2, the structural model demonstrated acceptable psychometric properties, including satisfactory convergent validity, composite reliability, and discriminant validity. Most importantly, the coefficient of determination for e-learning quality was R² = 0.712, indicating that the proposed dimensions and indicators explained 71.2% of the variance in e-learning quality. This finding reflects strong explanatory power and provides empirical support for the conceptual framework derived from the qualitative phase.

1. Table 2. Quantitative validation of the proposed model

Table note. These criteria were explicitly reported in the thesis as evidence of strong model validity.

The ANN analysis, presented in Table 3, clarified the relative importance of the indicators associated with blended learning and e-learning quality. Among all indicators, learner satisfaction obtained the highest optimal weight (0.2772), followed by learning effectiveness (0.1780). These findings indicate that the perceived quality of digital education within the proposed ecosystem depends most strongly on students’ positive educational experiences and the extent to which learning outcomes are effectively achieved. The ANN model also showed strong performance, with a reported MSE of 0.0060 and R² of 0.9580, suggesting a high level of predictive precision in estimating the importance of the indicators.

Finally, the MABAC ranking of AI capabilities is shown in Table 4. The results ranked intelligent support in the first position, followed by intelligent recommendation, automated assessment, and educational data analysis. This ranking indicates that, in the study context, the most influential AI capability for enhancing digital education quality is not merely analytical processing, but supportive and interactive functionality that directly assists learners throughout the learning process. These findings provide practical guidance for universities seeking to prioritize AI investments within blended learning environments.

1. Table 4. Ranking of AI capabilities using MABAC

Discussion

The findings support the central argument that improving digital education quality requires an integrated ecosystem rather than isolated technological interventions. The model designed in this study links pedagogy, technology, and educational quality in a single framework. This is consistent with the literature in the file that presents blended learning as a strategic educational design rather than a simple combination of modalities (Graham, 2009; Horn & Staker, 2014; Heinze & Procter, 2010; Istenič, 2024). The present findings extend that perspective by showing that blended learning becomes more effective when it is strengthened by AI capabilities that respond to learner needs in a targeted and adaptive manner. The prominence of flexibility, interaction, personalization, and infrastructure in the blended learning dimension suggests that quality digital education depends on both pedagogical design and institutional readiness. This aligns with the quality frameworks reviewed, including the Sloan-C approach and Quality Matters, both of which emphasize that successful online learning requires coherent design, support services, interaction, and accessibility. In the present study, these elements were not peripheral concerns; they formed the foundation upon which AI could operate meaningfully. Without adequate infrastructure and interaction, the benefits of intelligent systems would likely remain limited. The AI-related findings are also theoretically important. The ranking of intelligent support as the highest-priority capability suggests that the most valuable contribution of AI in higher education may be relational and facilitative rather than purely computational. This interpretation is compatible with the references in the file that discuss AI in terms of personalized assistance, adaptive learning, and enhanced student engagement (Hamadneh et al., 2022; Joseph et al., 2024; Kadhim & Hassan, 2020; Katsamakas et al., 2024). In other words, the study implies that AI is most educationally effective when it helps institutions humanize digital learning rather than merely automate it. This is a valuable insight for universities that may otherwise focus too heavily on analytics or administrative efficiency. The strong importance of learner satisfaction and learning effectiveness further reinforces the view that digital quality should be evaluated through both experiential and substantive outcomes. Satisfaction had the highest ANN weight, but it did not stand alone; it was closely followed by learning effectiveness, showing that a high-quality intelligent learning ecosystem must be both supportive and educationally productive. This result resonates with the quality models discussed, where learner experience, instructional effectiveness, and interaction are treated as interdependent dimensions. It also indicates that universities should avoid implementing AI merely as a technical innovation; instead, AI initiatives should be judged by whether they improve the real learning experience and outcomes of students. From an applied perspective, the study offers a practical roadmap for higher education institutions, especially Islamic Azad University. The research recommends investment in infrastructure development and faculty training, and the present findings make that recommendation even more compelling. Intelligent learning ecosystems require not only software and data, but also academic staff who can interpret, use, and integrate these tools into sound pedagogical practice. The findings therefore suggest that institutional strategy should begin with capacity-building, learner support systems, and coherent blended course design, while gradually expanding into more advanced AI-supported functions such as recommendation and automated assessment. At the same time, the discussion should acknowledge the limits of the study. The result itself indicates that the research was cross-sectional, which means that the model was validated at one point in time rather than across multiple stages of implementation. As a result, the findings provide a strong conceptual and empirical starting point, but further longitudinal and comparative studies would be helpful to examine how intelligent learning ecosystems evolve in practice. Even so, the present study makes an important contribution by offering a validated and context-sensitive framework that links blended learning, AI capabilities, and digital education quality in a way that is both theoretically meaningful and operationally useful.

Ethical Considerations

This study was conducted in accordance with the research ethics principles stated in the thesis and the research ethics charter of Islamic Azad University. Participation in both the interview and questionnaire phases was voluntary. Participants were informed about the purpose of the study, and confidentiality and anonymity were respected throughout data collection, analysis, and reporting. The researcher also adhered to the principles of originality, honesty in reporting, protection of participants’ information, and appropriate acknowledgment of all cited sources. A formal ethics approval code was not visible in the extracted file; therefore, none is reported here.

Conflict of Interest

The authors declare that there is no conflict of interest regarding the publication of this study.

Author Contributions

Fatemeh Akbari Markhali contributed to conceptualization, investigation, data collection, formal analysis, and original drafting of the manuscript. Mehdi Keramatpour contributed to supervision, methodology development, validation of the analytical framework, critical revision of the manuscript, and final approval of the version to be published. Both authors approved the final manuscript and agreed to be accountable for all aspects of the work.

Data Availability

This study was conducted in accordance with the research ethics principles stated in the thesis and the Research Ethics Charter of Islamic Azad University. Participation in the interview and questionnaire phases was voluntary, and all participants were informed about the purpose of the study. Confidentiality and anonymity were respected throughout data collection, analysis, and reporting. Given the focus of the study on artificial intelligence, blended learning, and digital education quality, the researcher also followed principles of responsible data use, academic honesty, protection of participants’ information, originality, and proper acknowledgment of all cited sources. As no formal ethics approval code was visible in the extracted file, no ethics code is reported here.

Funding

No specific external funding information was visible in the extracted file.

Acknowledgments

The authors would like to thank the experts, faculty members, and university staff who participated in the study and contributed to the development and validation of the proposed model.

References

Graham, Charles R. (2009). Blended learning models. In Encyclopedia of Information Science and Technology, Second Edition (pp. 375–382). IGI Global. https://doi.org/10.4018/978-1-60566-026-4.ch063

Halverson, Lisa R; Spring, Kristian J; Huyett, Sabrina; Henrie, Curtis R; & Graham, Charles R. (2023). Blended learning research in higher education and K-12 settings. In Learning, design, and technology: An international compendium of theory, research, practice, and policy (pp. 3107–3135). Springer. https://doi.org/10.1007/978-3-319-17461-7_31

Hamadneh, Nawaf N; Atawneh, Samer; Khan, Waqar A; Almejalli, Khaled A; & Alhomoud, Adeeb. (2022). Using artificial intelligence to predict students’ academic performance in blended learning. Sustainability, 14(18), 11642. https://doi.org/10.3390/su141811642

Heinze, Aleksej; & Procter, Chris. (2010). The significance of the reflective practitioner in blended learning. International Journal of Mobile and Blended Learning (IJMBL), 2(2), 18–29. https://doi.org/10.4018/jmbl.2010040102

Horn, Michael B; & Staker, Heather. (2014). Blended: Using disruptive innovation to improve schools. John Wiley & Sons.

Ilieva, Galina; Yankova, Tania; Klisarova-Belcheva, Stanislava; Dimitrov, Angel; Bratkov, Marin; & Angelov, Delian. (2023). Effects of Generative Chatbots in Higher Education. Information, 14(9), 492. https://doi.org/10.3390/info14090492

Istenič, Andreja. (2024). Blended learning in higher education: The integrated and distributed model and a thematic analysis. Discover Education, 3(1), 165. https://doi.org/10.1007/s44217-024-00239-y

Jones, Kevin Anthony; & Ravishankar, Sharma. (2021). Higher education 4.0: The digital transformation of classroom lectures to blended learning. Springer Nature. https://doi.org/10.1007/978-981-33-6683-1

Joseph, Sylvester; Tahir, Amna; Bibi, Farwa; Hamid, Khalid; Iqbal, Muhammad Waseem; Ruk, Sadaquat Ali; & Ahmad, Saleem Zubair. (2024). A Review Analysis on Using “AIED” to Improve Student Engagement in Hybrid Education. Bulletin of Business and Economics (BBE), 13(2), 424–435. https://doi.org/10.61506/01.00348

Kadhim, Mohammed K; & Hassan, Alia K. (2020). Towards Intelligent E-Learning Systems: A Hybrid Model for Predicatingthe Learning Continuity in Iraqi Higher Education. Webology, 17(2). https://doi.org/10.14704/web/v17i2/web17023

Katsamakas, Evangelos; Pavlov, Oleg V; & Saklad, Ryan. (2024). Artificial intelligence and the transformation of higher education institutions: A systems approach. Sustainability, 16(14). https://doi.org/10.3390/su16146118

Komsiyah, Indah. (2023). Control Quality of E-Learning Implementation Management. AL-ISHLAH: Jurnal Pendidikan, 15(2), 1881–1887. https://doi.org/10.35445/alishlah.v15i2.2972

Legon, Ron; & Garrett, Richard. (2018). The changing landscape of online education (CHLOE) 2: A deeper dive. Quality Matters & Eduventures Survey of Chief Online Officers.