City University of Hong Kong researchers unveil new Generative AI project on its collaborative ScholarMate platform, aiming to transform research and innovation processes globally, enhancing security and efficiency.
A trailblazing team of researchers from the City University of Hong Kong (CityUHK) is unveiling a revolutionary use of generative artificial intelligence (GenAI) technologies on its collaborative innovation network platform, ScholarMate. This project is poised to significantly transform science, technology and innovation (STI) services, fostering research and technology advancements.
Led by Ma Jian, a professor in the Department of Information Systems at CityUHK, the research team aims to address persistent issues in current STI services.
“By leveraging generative AI technologies on a collaborative innovation network platform, this project will help accelerate the pace of research and innovation,” Ma said in a news release.
A Decade of Progressive Development
Since its inception in 2007, ScholarMate has connected over 8 million users and accumulated 78 million research outputs.
The platform acts as a critical nexus, linking government funding bodies, universities, research institutes and technology companies, thereby promoting collaboration and sharing of expertise.
Building on this strong foundation, the latest project aims to digitally transform STI services. This transformation involves integrating cutting-edge federated learning technology with research knowledge graphs and large language models (LLMs).
Breaking Down Information Silos
The project seeks to resolve prevalent problems such as fragmented STI systems and information silos. Federated learning technology will be implemented to ensure data remains secure and personal privacy is preserved. This technology facilitates decentralized training of large language models, enhancing model accuracy while maintaining stringent security protocols.
Notably, the initiative will see the development of comprehensive research knowledge graphs. These graphs aim to map connections among researchers, projects, publications and patents, forming a detailed, domain-specific knowledge base.
Introducing STIGPT
A hallmark of this project is STIGPT (GPT for Science, Technology and Innovation), a generative AI model.
This model will provide intelligent assistant services through enhanced semantic search capabilities, intelligent recommendations and personalized content creation for research institutions and their personnel.
The Vision for the Future
The CityUHK team also envisions creating a low-code system development platform. This platform will enable institutions to implement digital transformation solutions with minimal programming effort, significantly lowering adoption barriers.
Integration of STIGPT with ScholarMate aims to construct “the first smart collaborative innovation network platform in the world that connects researchers from universities and industry to co-create IP knowledge,” leading to more streamlined processes and improved cooperation among stakeholders.
Commercialization and Market Potential
Recognizing the project’s immense market potential, the team, supported by the HK Tech 300 program, has established a startup named InnoCity.com. This venture focuses on harnessing GenAI technologies for various applications, including technology transfer and research marketing, effectively tapping into a promising market.