Lyuheng Yuan
CS PhD at Indiana University Bloomington
Email: lyyuan@iu.edu
[CV]
Embedded Files
Biography
Biography
My name is Lyuheng Yuan. I received my Ph.D. in Computer Science at Indiana University Bloomington in 2026, advised by Dr. Da Yan. My research focuses on graph and spatial data management, especially graph and spatial algorithms, parallel and distributed computing, GPU computing.
I received my master degree from University of Pennsylvania in 2021, and bachelor degree from Hefei University of Technology in 2019.
Research Interests
Research Interests
My research mainly focuses on designing and building efficient and scalable graph and spatial computing systems, which can be further split into following categories:
- Parallel and Distributed Subgraph Mining
Frequent Subgraph Mining: Subgraph mining is a family of problems which tries to find desirable patterns from a large graph. In particular, my work has focused on frequent subgraph pattern mining (FSM).
Dense Community Detection: Identifying dense communities in networks has important applications in social network recommendation and uncovering significant biomolecular structures and pathways.
Demostration: We developed a series of user-friendly graphical interface that allows users to intuitively control and adjust input parameters.
- Graph Computing on Modern Hardware
Subgraph Matching: Graph mining problems are usually computing-intensive which can be prohibitively time-consuming using conventional methods. Therefore, we have developed parallel and distributed approaches aiming to boost its efficiency.
Subgraph Matching: Graph mining problems are usually computing-intensive which can be prohibitively time-consuming using conventional methods. Therefore, we have developed parallel and distributed approaches aiming to boost its efficiency.
DFS on GPU: ICDE'24
General-Purpose Framework: We aim to develop a general and unified framework capable of mining a wide range of subgraph patterns. This allows users to accomplish their tasks through our interfaces without needing to worry about low-level programming details.
Adaptive BFS framework on GPU: ICDE'24
General parallel framework on CPU: TKDE'25
SPARQL query on RDF data: BigData'25
Graph Analytics: Iterative graph traversal, where memory I/O becomes the dominant overhead.
k-core: ICDE'23
- Parallel and Distributed Spatial Computing
Spatial data generated by satellites and GPS are often massive in size and include various irregular object shapes, such as points, lines and polygons which are complex to handle. Besides, answering spatial queries such as finding intersections, containment and nearest neighbors are computing expensive, especially over large datasets.
Page updated
Google Sites
Report abuse