HPGDML 2017 : High Performance Graph Data Mining and Machine Learning Workshop
|Dates||2017/11/18 (iCal) - 2017/11/18|
|Location:||New Orleans, USA|
|Table of Contents|
HPGDML'17: High Performance Graph Data Mining and Machine Learning Workshop
Held in conjunction with IEEE ICDM 2017: 17th IEEE International Conference on Data Mining (IEEE ICDM 2017)
Workshop website: http://hpgdml17.bsc.es
Call for Papers
Applications which need to manage and process large scale graph data have become prominent in recent times. Social network analysis, semantic web, bioinformatics, cheminformatics, etc. are some examples for such application domains which deal with large graphs of millions and billions of vertices. Graph processing have attracted significant attention from High Performance Computing (HPC) community due to complexities involved with processing and storage of large graphs. Significant body of research have been conducted in recent times to address the void of large graph data mining. New programming models such as Pregel, graph processing frameworks such as Giraph, Hama, and libraries such as GraphLab, PBGL, ScaleGraph have been developed to address the need of software for high performance large graph processing. Furthermore, large scale distributed memory compute clusters, single shared memory high performance computers, heterogeneous hardware such as GPGPUs, FPGAs have been tested for carrying out large graph data processing tasks. These efforts have been bolstered by graph related benchmarking initiatives such as Graph 500, Green Graph 500, etc. Despite these significant research efforts, there exist significant issues and technical gaps which need to be solved in the area of high performance graph data mining. On the other hand in recent times we observe significant rise of the research conducted on the intersection of graph mining and machine learning. Rise of the deep learning techniques have brought machine learning to a new level.
High Performance Graph Data Mining and Machine Learning 2017 (HPGDML’17) workshop aims to provide a unified platform for discussing the latest state-of-the-art efforts conducted to address such research issues related to high performance large graph data mining and machine learning. HPGDML’17 will be held in conjunction with the IEEE International Conference on Data Mining (ICDM) 2017. The workshop will take place in the Roosevelt New Orleans, New Orleans, USA on November 18th 2017. We invite researchers from academia and industry who work in graph data mining and machine learning in high performance computing environments to submit their original (full/short) papers. Submissions will be peer reviewed in single-blinded manner and each submission will receive minimum three reviewer comments.
- Novel large graph data management systems
- Deep Learning and its applications
- Novel large graph processing frameworks and programming paradigms
- Graph processing in many core processors such as GPGPUs/FPGAs, Xeon Phi, etc.
- Graph data mining in HPC Clouds
- Workflows which involve both graph data mining and machine learning
- HPC graph databases and query languages
- Novel graph partitioning algorithms
- Application experiences of large graph processing on HPC environments
- Benchmarks for large graph processing workloads
- Performance characterization of large graph mining tasks
- Scalable graph analysis algorithms and novel data structures
- High performance streaming graph processing algorithms
A submitted paper should be of one of two types:
1. Regular Research Paper: The paper will report original research results with sound evaluation. It should be at most 8 pages.
2. Short Paper: The paper will present an on-going work or a survey. It will clearly state the problem to be addressed, an outline of the methodology that the authors plan to follow, and preliminary evaluation results. It should be at most 4 pages.
Only electronic submissions in PDF format will be considered. Submitted manuscripts may not exceed the specified page limits including tables, figures, and references. All submissions must be prepared formatted according to IEEE ICDM format requirement. The papers must be formatted as single-blinded. Submissions must be uploaded to the below mentioned submission site.
The proceedings of the HPGDML'17 will be included in the IEEE ICDM 2017 Workshops Proceedings volume published by IEEE Computer Society Press, and will also be included in the IEEE Xplore Digital Library.
Paper submission deadline (extended): August 1st 2017, AoE Decision notification: September 4th 2017, AoE Camera-ready deadline: September 15th 2017, AoE Workshop date: November 18th 2017, AoE
General Co-Chairs: Toyotaro Suzumura (IBM T.J. Watson Research Center, USA) Dario Garcia-Gasulla (Barcelona Supercomputing Center, Spain)
Program Committee Chair: Miyuru Dayarathna (WSO2, Inc., USA)
Leman Akoglu (Carnegie Mellon University, USA) Lijun Chang (University of New South Wales, Australia) Sameh Elnikety (Microsoft Research, USA) Stephan Günnemann (Technical University of Munich, Germany) Xutao Li (Harbin Institute of Technology, China) Makoto Onizuka (Osaka University, Japan) Roger Pearce (LLNL, USA) René Peinl (Hof University, Germany) Arnau Prat (Sparsity-Technologies, Spain) Jason Riedy (Georgia Institute of Technology, USA) Amitabha Roy (Intel Labs, USA) Sherif Sakr (University of New South Wales, Australia) Erik Saule (University of North Carolina at Charlotte, USA) Julian Shun (University of California, Berkeley, USA) Hang Hang Tong (Arizona State University, USA) Yinglong Xia (Huawei Research America, USA)
In case of questions, please contact the Program Committee Chair via miyurud at wso2.com .
|Acronym||HPGDML 2017 +|
|End date||November 18, 2017 +|
|Has coordinates||29° 57' 0", -90° 4' 12"Latitude: 29.949933333333|
Longitude: -90.070116666667 +
|Has location city||New Orleans +|
|Has location country||Category:USA +|
|Start date||November 18, 2017 +|
|Submission deadline||August 1, 2017 +|
|Title||HPGDML 2017 : High Performance Graph Data Mining and Machine Learning Workshop +|