Jinan University Pioneers Interdisciplinary AI Graduate Training Model
In the rapidly evolving landscape of artificial intelligence (AI), the demand for highly skilled, interdisciplinary talent has never been greater. As nations race to secure leadership in AI innovation, the role of higher education institutions in cultivating the next generation of researchers and engineers has become paramount. Against this backdrop, a team of researchers from the University of Jinan has introduced a groundbreaking model for graduate education in AI, one that seamlessly integrates multiple disciplines, strengthens industry-academia collaboration, and redefines the framework for training future technological leaders.
Published in the December 2021 issue of Software Guide, Han Shi-yuan, Chen Yue-hui, Wu Peng, Hao Xiao-yan, and Zhou Jin present a comprehensive and empirically validated approach to interdisciplinary AI graduate training. Their work, titled “Research and Practice of Postgraduate Talent Training in Artificial Intelligence under Multi-Disciplinary Cross,” outlines a transformative educational framework that has already demonstrated significant success in cultivating high-caliber AI professionals. The model, developed and implemented at the University of Jinan’s Artificial Intelligence Research Institute, is not merely a theoretical construct but a living, evolving system that has produced tangible outcomes in research, innovation, and student achievement.
The urgency of this initiative is rooted in both national strategy and regional economic development. In January 2020, China’s Ministry of Science and Technology officially approved Jinan as a national pilot zone for the development of next-generation AI technologies. This designation placed a significant responsibility on local academic institutions to align their educational and research missions with the broader goals of technological sovereignty and industrial transformation. The University of Jinan, already recognized as a provincial “high-level university,” was uniquely positioned to respond to this call. However, as the authors note, the traditional academic structure, with its rigid departmental boundaries, posed a formidable barrier to the kind of fluid, cross-domain collaboration that AI inherently requires.
Recognizing this challenge, the research team set out to dismantle the silos that often hinder innovation in higher education. Their solution was not a minor adjustment to an existing curriculum but a complete reimagining of the graduate training process. The result is a model that is both ambitious in scope and meticulous in execution, built on five core pillars: strategic top-level design, a precisely coupled interdisciplinary curriculum, a robust system of process-based assessment, deep integration with industry, and a novel approach to student management through a grid-based system.
At the heart of the model is a clear and forward-looking vision for what an AI graduate should become. The program is designed to produce not just experts in algorithms or machine learning, but versatile engineers and researchers capable of applying AI to complex, real-world problems in fields as diverse as logistics, manufacturing, healthcare, and agriculture. This vision is reflected in the program’s five designated research directions: intelligent computing theory and application, smart logistics and supply chain management, intelligent control of complex systems, intelligent manufacturing and smart agricultural machinery, and bioinformatics and intelligent healthcare. Each of these areas represents a convergence of AI with another discipline, ensuring that students are trained to think and work at the intersection of multiple domains.
The curriculum itself is a masterclass in educational engineering. Drawing on the CDIO (Conceive, Design, Implement, Operate) framework, which emphasizes hands-on, project-based learning, the team has structured the program into three distinct but interconnected modules. The first, the AI Foundation module, provides students with the essential mathematical and theoretical underpinnings of the field. Courses such as “Engineering Matrix Theory,” “Numerical Analysis,” and “Artificial Intelligence and Machine Learning” ensure that all students, regardless of their background, have a solid grounding in the core principles of AI.
The second module, the Interdisciplinary Direction module, is where the program truly distinguishes itself. Instead of offering a generic set of electives, the curriculum is tailored to the five research directions, with courses taught by faculty from across the university. For instance, students focusing on smart logistics take classes in supply chain modeling and simulation from the Business School, while those in intelligent healthcare study bioinformatics and protein structure prediction with faculty from the College of Biological Science and Technology. This cross-college collaboration is not just a logistical arrangement; it is a deliberate strategy to expose students to the language, methods, and challenges of other disciplines, fostering a mindset of interdisciplinary inquiry.
The third module, Practice and Literacy Development, is designed to cultivate the soft skills and professional competencies that are essential for success in the modern workforce. Courses in ethics, leadership, and entrepreneurship are delivered through a series of lectures by leading academics, industry experts, and entrepreneurs. Additionally, students engage in “AI Engineering Comprehensive Training,” a capstone course where they work on real-world case studies, often drawn from industry partnerships. This experiential learning component is critical, as it allows students to apply their knowledge in a practical context, bridging the gap between theory and application.
Equally important is the model’s emphasis on accountability and continuous improvement. The authors recognize that innovation in education cannot be left to chance; it must be supported by a rigorous system of quality assurance. To this end, they have implemented a comprehensive process assessment framework that monitors student progress at every stage of their academic journey. Key milestones such as the thesis proposal, mid-term review, and final defense are subject to strict evaluation criteria, with a “bottom-elimination” mechanism in place to ensure that only students who meet the highest standards advance. This may sound harsh, but the authors argue that it is necessary to maintain the integrity of the program and to prepare students for the competitive realities of the professional world.
At the same time, the model is not purely punitive; it also includes a range of incentives to motivate and reward excellence. Students who publish in high-impact journals, win national competitions, or secure patents are recognized and supported, with additional funding and opportunities for international exchange. The program has even developed its own classification system for research outputs, aligning with the prestigious CCF (China Computer Federation) recommendations to ensure that achievements are evaluated by the highest academic standards. This dual approach of “carrot and stick” creates a culture of high expectations and continuous striving, where students are both challenged and supported.
Perhaps the most innovative aspect of the model is its deep integration with industry. The authors understand that AI is not an academic exercise; it is a technology that must be tested, refined, and deployed in the real world. To facilitate this, the University of Jinan has established a series of joint training bases with leading technology companies such as Inspur Group and Zhongchuang Software. These partnerships are not superficial; they are embedded into the fabric of the program. Students spend significant time at these sites, working on real projects under the guidance of both academic and industry mentors. This “dual-advisor” system ensures that students receive guidance from both theoretical and practical perspectives, a balance that is often missing in traditional graduate programs.
The industry collaboration extends beyond internships. The program actively incorporates industry case studies into its curriculum, ensuring that classroom learning is always connected to real-world challenges. Moreover, the research conducted by students and faculty is often directly applicable to the needs of partner companies, creating a virtuous cycle of innovation and commercialization. This close alignment with industry not only enhances the relevance of the program but also increases the employability of its graduates, many of whom are recruited directly by the companies they worked with during their studies.
Another distinctive feature of the model is its approach to student management. Rather than treating graduate students as isolated individuals, the program adopts a “grid management” system that organizes them into a network of support and accountability. This system is based on both horizontal and vertical structures: horizontally, students are grouped by their research areas and physical locations (such as labs and dormitories); vertically, they are supported by a hierarchy of advisors, teaching assistants, and student organizations. This dual-axis structure ensures that no student falls through the cracks, and that issues related to academic progress, mental health, or personal well-being are promptly identified and addressed.
The effectiveness of this model is not merely anecdotal; it is backed by a wealth of data. Since its implementation in 2018, the program has trained over 150 AI-focused graduate students. It has produced four provincial-level outstanding master’s theses, a remarkable achievement that speaks to the quality of research being conducted. Students have also won five national-level discipline competitions, demonstrating their ability to compete at the highest levels. In terms of curriculum development, the program has established eight provincial-level high-quality courses and two provincial graduate training bases, setting a new standard for AI education in regional universities.
The impact of this work extends far beyond the walls of the University of Jinan. As other institutions grapple with the challenges of interdisciplinary education, this model offers a clear and actionable roadmap. It demonstrates that with the right vision, leadership, and institutional support, it is possible to break down the barriers between disciplines and create a truly integrated learning environment. It also shows that industry collaboration is not a luxury but a necessity, especially in a field as dynamic and application-driven as AI.
Moreover, the model has important implications for the broader landscape of higher education in China. While elite universities in Beijing and Shanghai often dominate the national conversation, this work highlights the vital role that provincial universities can play in driving innovation and serving regional economies. The University of Jinan’s success is a testament to the fact that with strategic investment and bold thinking, even non-elite institutions can become leaders in cutting-edge fields.
The authors are careful to note that their model is not a one-size-fits-all solution. Different universities have different strengths, resources, and regional contexts. What works in Jinan may not work elsewhere. However, the core principles—interdisciplinarity, industry integration, rigorous assessment, and student-centered management—are universally applicable. They represent a new paradigm for graduate education, one that is responsive to the demands of the 21st century.
As AI continues to reshape every aspect of society, the need for well-trained, ethically grounded, and innovative professionals will only grow. The model developed by Han Shi-yuan, Chen Yue-hui, Wu Peng, Hao Xiao-yan, and Zhou Jin at the University of Jinan offers a powerful example of how higher education can rise to this challenge. It is not just a training program; it is a blueprint for the future of AI education.
Han Shi-yuan, Chen Yue-hui, Wu Peng, Hao Xiao-yan, Zhou Jin, University of Jinan, Software Guide, DOI:10.11907/rjdk.212257