Dr. Songjie Niu

Postdoc

Tsinghua University

I am a postdoc at Department of Computer Science, Tsinghua University. My research interests are in graph data processing, artificial intelligence, and data management systems. I received my Ph.D. in Computer Science from Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS), University of Chinese Academy of Sciences (UCAS), and my B.E. from Beijing Institute of Technology.

 

During my Ph.D., I conducted in-depth research on performance modeling and optimization of large-scale distributed in-memory graph computation. My work includes performance analysis and modeling of graph computation, optimization of complex representative graph algorithms like second-order random walk, and graph computation optimizations combining storage hardware such as CPU caches and NVM persistence. I developed a lightweight general-purpose distributed in-memory graph computation platform named GraphLite, which is open-sourced on GitHub and has been used by over 1,000 users.

 

During my postdoctoral period, I have been engaged in research on the integration of graph data processing and large language models. I am currently participating in the AI4EDU project at Tsinghua University. Knowledge graphs can provide a structured knowledge foundation for large language models. By integrating knowledge graphs with large language models, I aim to help the AI4EDU system better understand and respond to student needs, providing more accurate and context-aware educational content.