DKE Lab

Welcome!

How do you do?
This is Data & Knowledge Engineering Lab chair, Lee, Young-Koo.

Cultivating leaders of challenge and exploration that will change the world is both DKE Lab's historical mission and its pride.

Who we are?

Welcome to DKE Lab, where innovation thrives and boundaries are challenged. We are a dynamic team of researchers dedicated to advancing knowledge and solving real-world problems through interdisciplinary collaboration.

At DKE Lab, our passion lies in exploring a diverse array of fields, including Big-Data Analytics, Distributed & Cloud Computing, Advanced Deep Learning, Graph Neural Networks, Database & Query Optimization, and Graph Augmentation. Through our specialized expertise and relentless pursuit of excellence, we strive to push the frontiers of technology and make meaningful contributions to our respective fields and beyond.

Driven by a shared vision of scientific exploration, our team brings together individuals with diverse backgrounds and expertise. We believe in the power of collaboration and recognize that tackling the most pressing challenges facing society today requires a multidisciplinary approach. Through rigorous research, experimentation, and collaboration, we aim to unravel the complexities of our world and develop innovative solutions that drive positive change.

Whether it's delving into the depths of data, harnessing the scalability of distributed systems, or pushing the boundaries of artificial intelligence, we are committed to pushing the boundaries of what's possible and shaping the future of technology.

Our Research Areas

Distributed Frameworks Development

Design and development of distributed frameworks for various applications including Graph Neural Networks and Big Data.

Big-Data Analytics

Employing advanced techniques to extract valuable insights from vast and diverse datasets, empowering informed decision-making.

Distributed & Cloud Computing

Leveraging distributed systems and cloud infrastructure to enable seamless scalability, fault tolerance, and resource optimization.

Advanced Machine Learning and AI

Delving into complex neural network architectures and methodologies to solve complex problems across various domains.

Graph Neural Networks Toolset

Toolsets for Graph Neural Network architectures training and inference, including distributed framework and graph augmentation.

Query Optimization

Streamlining database operations and query processing to enhance performance and optimize resource utilization.

What we are doing?

Artificial Intelligence Innovation Hub Conducting challenging research and development to solve technical and scientific challenges in the AI field based on an open joint research system. - By challenging high-risk and challenging challenges to break through the technical limitations of the AI field, it secures the world's best original technology in the AI field. - Contributing to the improvement of human quality of life by developing AI technology to solve scientific challenges with great economic and social ripple effects. - Universal intelligence machines that can be applied universally regardless of domain to lead the future AI industry.

Development of a distributed graph DBMS for intelligent processing of big graphs Developing a distributed GDBMS for intelligent high-speed processing of ultra-large graphs. In this task, 1) an intelligent graph application in which three main graph queries (pattern search, analysis, and learning questions) are mixed can be developed on one system, 2) and 3) a cloud-based distributed GDBMS that provides scalable and efficient high-speed query processing for ultra-large graphs is designed and developed. Open source the GDBMS technologies to be developed so that they can be widely used across industries related to graph applications. - Aim to secure the following core technologies to exceed world-class RFP requirements. - Processing distributed disk-based large/dynamic graph queries. - Processing high-speed queries considering worst-optimal join. - Query optimization techniques considering multiple join algorithms and binary join algorithms. - Integrated development technology of three queries using an integrated API library. - Giant graph neural network learning and inference techniques. - RDMA-based high-speed network manager technology. - By securing core technologies, we aim to improve the following six key performance figures by more than 10% to 20% above the RFP target figure, ultimately securing world-class technology. - Improved performance of megaprograph analysis by more than 10% (PageRank: 12 seconds, Triangle Counting: 45 seconds). - Leverage limited computing resources to increase the analytical graph size by more than 50% (1.5 trillion edges). - More than 15% storage throughput per second for time-varying graph data (0.7 billion edges). - More than 10% improvement in incremental processing performance over static processing of graphs (PageRank: 45%, Label Propagation: 12%). - More than 10% improvement in fast graph machine learning processing speed (GCN-1: 0.27 s, GCN-2: 0.72 s, and GCN-3: 0.9 s). - Reduce performance differences by more than 10% between SSD-based query processing and pure memory-based processing (within 4.5%).

Summarization and Compression based Fast RDF Query Processing Techniques for Massive RDF Graphs In this study, large-capacity RDF graphs are effectively summarized/compressed, and high-speed query processing technology is developed from compressed RDF graphs. And, the RDF graph summary/compression technology is extended to distributed processing. It develops summary and compression graph processing tools and APIs to process large-capacity RDF graphs

Bio-Mechatronics Practical problem research group The Bio-Mechanics Practical Problems Research Group will establish a steering committee headed by the general manager as its vision for self-directed creative talent development according to the development of core technology specialization and convergence technology. It will provide an opportunity for education and experience to derive specific empirical practical problems in connection with the industry and cultivate the ability to creatively solve them. Based on this, we intend to continue to promote talent development projects by establishing a field-specific engineering talent development support system as a knowledge-sharing system within an intelligent industry-academic cooperation platform.

Our Achievements

Publications Awards

Journal Sultana, Tangina, Md Delowar Hossain, Md Golam Morshed, Tariq Habib Afridi, and Young-Koo Lee. "Inductive autoencoder for efficiently compressing RDF graphs." Information Sciences 662 (2024): 120210.
Journal Rahman, Md Arif, and Young-Koo Lee. "Forester: Approximate Processing of an Imperative Procedure for Query-Time Exploratory Data Analysis in a Relational Database." Electronics 13, no. 4 (2024): 759.
Conference Morshed, Md Golam, and Young-Koo Lee. "LFP: Layer Wise Feature Perturbation based Graph Neural Network for Link Prediction." In 2023 IEEE International Conference on Big Data and Smart Computing (BigComp), pp. 85-91. IEEE, 2023.
Patent Jeong, Taeyoung, P. A. R. K. Kisung, Chanho Jeong, Ki Hong Kim, Young-Koo Lee, and Md Mostofa Kamal Rasel. "Operator movement optimization technique for procedures." U.S. Patent Application 17/561,104, filed February 16, 2023.

View All

BigComp 2022 Md Golam Morshed won the Best Paper Award (3rd Place) at BigComp 2022 for his paper titled "Mnssd: A real-time dnn based companion image data annotation using mobilenet and single shot multibox detector."
KDBC 2020 Tangina Sultana, Delowar Hossain, and Golam Morshed won the Best Paper Award at KDBC 2023 for their paper titled "Enhancing Node Classification in Graphs through Mapping-Based Contrastive Learning."
KDBC 2020 Aftab Alam, Numan Khan, Tariq Afridi, and Tangina Sultana won the Best Paper Award at KDBC 2020 for their paper titled "SCBVR: A Hybrid Approach for Large-scale Video Retrieval."
KDBC 2020 Tangina Sultana, Mostafa Kamal Rasel, and Aftab Alam won the Best Paper Award at KDBC 2020 for their paper titled "Frequent and Semantically Related Pattern Detection Based Encoding and Compression for Knowledge Graphs."
KCC 2020 Numan Khan, Aftab Alam, Md Anwarul Islam, and Jawad Khan won the Best Paper Award at KCC 2020 for their paper titled "DISIVR: distributed deep feature indexer for video big dataretrieval on spark."

Professor Lee, Young-Koo

Professor Lee Young-Koo is a distinguished figure in the realm of computer science, renowned for his groundbreaking contributions to academia and research. Currently serving as a Professor at Kyung Hee University (Global Campus), South Korea, Professor Lee Young-Koo brings a wealth of expertise and leadership to their role.

Present
March 2015
Professor
Kyung Hee University (Global Campus), South Korea
February 2015
March 2010
Associate Professor
Kyung Hee University (Global Campus), South Korea
February 2010
March 2004
Assistant Professor
Kyung Hee University (Global Campus), South Korea
February 2004
September 2002
Postdoctoral Research Associate
University of Illinois at Urbana-Champaign
February 2004
March 2002
Post Doctoral Fellow
Advanced Information Technology Research Center(AITrc), KAIST, South Korea
February 2002
March 1994
PhD in Computer Science
Korea Advanced Institute of Science and Technology, South Korea
February 1994
March 1992
Master's in Computer Science
Korea Advanced Institute of Science and Technology, South Korea
February 1992
March 1988
Bachelor's in Computer Science
Korea Advanced Institute of Science and Technology, South Korea