SAN DIEGO, September 28, 2021 / PRNewswire / – The Association for Computing Machinery (ACM) Special Interest Group on Knowledge Discovery and Data Mining (SIGKDD) today announced the recipients of the SIGKDD Best Paper Awards, honoring the papers presented at the SIGKDD Annual Conference that advance fundamental understanding of the field of data knowledge discovery and data mining. The winners were selected from over 2,200 papers initially submitted for review for presentation at KDD 2021, which took place August 14-18. Of the 394 papers chosen for the conference, three prizes were awarded: best research paper, best applied data science paper and best student paper.
âAcademic and industrial researchers from around the world have submitted papers to KDD 2021 to showcase the latest innovations in the field of knowledge discovery in machine learning,â noted Dr Haixun Wang, chair of the SIGKDD Prize committee. âThose who were selected for recognition have pushed the boundaries of machine learning, especially for solving real-world problems. The best SIGKDD articles of 2021 are:
- Research Track: “Fast and Efficient Tucker Decomposition in Memory to Meet Diversified Time Range Queries”, by Jun-gi Jang and U Kang (both from Seoul National university) – After studying methods that analyze dense tensors to discover hidden factors, the researchers showed that the Zoom-Tucker is a fast and efficient in-memory Tucker decomposition method to find hidden factors of time tensor data in a arbitrary time range. The article shows that by carefully decoupling the preprocessed results included in various time ranges and carefully determining the order of the calculations, the Zoom-Tucker method is up to 171.9 times faster and requires up to 230 times less space than existing methods, offering a creative solution that yields amazing results.
- Research path, student article: “Spectral grouping of attributed multi-relational graphs”, by Ylli Sadikaj (University of Vienna), Yllka Velaj (University of Vienna), Sahar Behzadi Soheil (University of Vienna), and Claudia plant (University of Vienna) – After studying the challenge of grouping graphs when complex data in many domains is represented as both attributed and multi-relational networks, the researchers proposed SpectralMix, a joint dimensionality reduction technique for multi-graphs -relational with categorical node attributes. SpectralMix integrates all the information available from the attributes, the different types of relationships and the structure of the graph to allow a good interpretation of the clustering results.
- Applied Data Science Course: “Supporting the policy response to COVID-19 with large-scale mobility-based modeling”, by Serina chang (Stanford University), Mandy wilson (University of Virginia), Bryan leroy lewis (University of Virginia), Zakaria Mehrab (University of Virginia), Emma J. Pierson (Microsoft Research), Pang Wei Koh (Stanford University), Jaline GÃ©rardin (Northwestern University), Beth red bird (Northwestern University), David Grusky (Stanford University), Madhav Marathi (University of Virginia), and Jure Lesovec (Stanford University) – The authors have introduced a decision support tool that uses large-scale data and epidemiological modeling to quantify the impact of mobility changes on infection rates. The model captured the spread of COVID-19 using a dynamic, fine-grained mobility network that encodes the hourly movements of people from neighborhoods to individual locations, with more than 3 billion hourly edges. The paper describes the robust IT infrastructure required to support millions of model builds that can simulate a wide variety of reopening plans, providing decision makers with an analytical tool to assess the trade-offs between future infections and mobility restrictions.
- Applied data science course, finalist: “Energy-efficient 3D vehicular crowdsourcing for disaster response through distributed deep reinforcement learning”, by Hao Wang (Beijing Institute of Technology), Chi (Harold) Liu (Beijing Institute of Technology), Zipeng Dai (Beijing Institute of Technology), Jian tang (DiDi Chuxing), and Guoren Wang (Beijing Institute of Technology) – The authors introduced DRL-DisasterVC (3D), a distributed deep reinforcement learning framework, to maximize the amount of data collected from unmanned vehicles in a work area in the event disaster relief in 3 dimensions (3D). The article described a 3D convolutional neural network with multi-head relational attention for spatial modeling and control of auxiliary pixels for space exploration, as well as a new disaster response simulator, called “DisasterSim”, used for conduct extensive experiments to show that DRL-DisasterVC (3D) maximizes data collection, geographic equity, and energy efficiency, while minimizing data loss due to limited transmission rate.
The Technical Program Committees for the Research Track and the Applied Data Science Track identified and nominated a very selective group of papers for the Best Paper Awards. The nominated papers were then independently reviewed by a committee headed by Chairman Haixun Wang, vice president of engineering and algorithms at Instacart; Professor Wei wang, University of California, Los Angeles; Professor Beng chin, national university of singapore; Professor Jiawei Han, University of Illinois at Urbana-Champaign; and Sanjay Chawla, Research Director of the Data Analysis Department of Qatar Computing Research Institute.
For more information on KDD 2021, please visit: https://www.kdd.org/kdd2021/.
About ACM SIGKDD:
ACM is the world’s premier professional organization for researchers and professionals dedicated to advancing the science and practice of knowledge discovery and data mining. SIGKDD is ACM’s Special Interest Group on Knowledge Discovery and Data Mining. The annual KDD International Conference on Knowledge Discovery and Data Mining is the premier interdisciplinary conference for data mining, data science, and analytics.
Facebook – https://www.facebook.com/SIGKDD
LinkedIn â https: //www.linkedin.com/groups/160888/
SOURCE ACM SIGKDD