The current research ecosystem faces a major structural challenge: while massive amounts of research—over 3 million papers annually—are produced, the knowledge often remains siloed, failing to reach business executives, policymakers, or even other academics who need it. This disconnect hinders innovation, manifesting as “technology information silos” that complicate writing, publication, and citation for scholars, and delay access to actionable insights for industry.
Beyond Knowledge Storage
Since 2007, City University of Hong Kong’s College of Business has led the development of ScholarMate.com, which has become mainland China’s largest professional research social network platform. This platform serves as a critical connector, linking government funding agencies (such as the National Natural Science Foundation of China), hundreds of universities, research institutions, and thousands of technology companies. It enables the structured sharing and intelligent matching of research outcomes, including papers, patents, and projects.
Building on this, the startup InnoCity was established. It is developing STIGPT, a vertical domain large language model. This system uses the platform’s accumulated data to construct dynamic research knowledge graphs, accurately mapping semantic relationships between research outputs, personnel, and institutions. By strictly adhering to data security and privacy, it is building professional knowledge graphs and vector knowledge bases, shifting the focus from mere “knowledge sharing “ to “knowledge co-creation.” This integration of platform accumulation with Generative AI forms the core of Social AI for Research. It is more than an AI tool application; it is a reconstruction of the research social network to enable understandable, interactive, and transformative knowledge circulation across academia, industry, and government.
From Solitary Research to Intelligent Collaboration
Traditional research is often a “solitary climb,” characterised by long cycles and limited efficiency as researchers independently handle literature reviews, data cleaning, and revisions. In the current data explosion, this solitary model is unsustainable.
The value of AI lies in its function as a well-informed and tireless research partner. For example, if a researcher inputs a topic into STIGPT, the system can instantly:
1. Map the global research landscape.
2. Identify theoretical gaps, e.g., The Role of Employee Psychological Safety in AI Transformation.
3. Suggest local case studies relevant to industrial policies, e.g., in the Greater Bay Area (GBA).
4. Optimise language expression to suit target journals, e.g., Decision Support Systems.
5. Simulate peer review through the Peer Digital Scholar feature, which users have reported leads to significantly higher paper acceptance rates.
After publication, research is transformed into high-dimensional knowledge vectors for precise promotion. This facilitates the rapid transformation of academic value into tangible commercial value. For instance, a study on cross-border data flows was adopted by a Shenzhen tech company to optimise its overseas R&D centre layout.
Connecting Academia and Industry in the Greater Bay Area
For the Guangdong-Hong Kong-Macao Greater Bay Area (GBA), which aims to be an “International Centre for Science, Technology and Innovation,” Social AI is crucial. True innovation in the GBA stems from the efficient circulation of knowledge, talent, and need: universities propose theories, enterprises convert them, and emerging challenges are fed back to academia, forming a virtuous cycle.
ScholarMate serves as the infrastructure for this cycle, hosting dedicated pages for GBA institutions such as CityUHK, Sun Yat-sen University, Huawei, Tencent, and BYD. The platform enables two-way interaction:
Enterprises to Academia: Business executives can post real-world challenges to invite “problem-oriented” research.
Academia to Enterprises: Managers can search for cutting-edge academic papers, which the system automatically distills into concise management recommendation reports. Users can then directly engage with the paper’s author through their “Digital Scholar” for immediate dialogue.
This breaks down the barriers between the “academic ivory tower” and “industrial practice.” The platform is evolving from an academic assisting tool to a smart engine for regional innovation governance, as seen in its collaboration with the Shenzhen Nanshan District Government to provide data-driven policy-making support.
Building the Future Intelligent Research Ecosystem
The development of the ScholarMate platform aligns with City University of Hong Kong’s mission to serve the Greater Bay Area. The platform invites:
・ Business School Faculty to use it as an intelligent assistant for improving high-tier journal publication efficiency.
・ Graduate Students to leverage it as a research accelerator for efficient output and expanded academic influence.
・ GBA Business Executives to utilise it as an “external brain” connecting frontier academia with business practice for innovative decision-making.
Social AI for Research, built on a decade of platform accumulation and breakthroughs in STIGPT, is opening up new possibilities. The future of research is not solitary but collaborative, with innovation emerging from the close interaction between universities, enterprises, and society.
ScholarMate—Making research more successful, innovation more efficient.
Now available for trial use by universities, research institutions, and enterprises in Hong Kong and the Greater Bay Area.
