Publication detail

Towards a Novel Infrastructure for Conducting High Productive Cloud-Based Scientific Analytics

BREZANY, P. LUDESCHER, T. FEILHAUER, T.

Original Title

Towards a Novel Infrastructure for Conducting High Productive Cloud-Based Scientific Analytics

English Title

Towards a Novel Infrastructure for Conducting High Productive Cloud-Based Scientific Analytics

Type

conference paper

Language

en

Original Abstract

The life-science and health care research environments offer an abundance of new opportunities for improvement of their efficiency and productivity using big data in collaborative research processes. A key component of this development is e-Science analytics, which is typically supported by Cloud computing nowadays. However, the state-of-the-art Cloud technology does not provide an appropriate support for high-productivity e-Science analytics. In this paper, we show how productivity of Cloud-based analytics systems can be increased by (a) supporting researchers with integrating multiple problem solving environments into the life cycle of data analysis, (b) parallel code execution on top of multiple cores or computing machines, (c) enabling safe inclusion of sensitive datasets into analytical processes through improved security mechanisms, (d) introducing scientific dataspace-a novel data management abstraction, and (e) automatic analysis services enabling a faster discovery of scientific insights and providing hints to detect potential new topics of interests. Moreover, an appropriate formal productivity model for evaluating infrastructure design decisions was developed. The result of the realization of this vision, a key contribution of this effort, is called the High-Productivity Framework that was tested and evaluated using real life-science application domain addressing breath gas analysis applied e.g. in the cancer treatment.

English abstract

The life-science and health care research environments offer an abundance of new opportunities for improvement of their efficiency and productivity using big data in collaborative research processes. A key component of this development is e-Science analytics, which is typically supported by Cloud computing nowadays. However, the state-of-the-art Cloud technology does not provide an appropriate support for high-productivity e-Science analytics. In this paper, we show how productivity of Cloud-based analytics systems can be increased by (a) supporting researchers with integrating multiple problem solving environments into the life cycle of data analysis, (b) parallel code execution on top of multiple cores or computing machines, (c) enabling safe inclusion of sensitive datasets into analytical processes through improved security mechanisms, (d) introducing scientific dataspace-a novel data management abstraction, and (e) automatic analysis services enabling a faster discovery of scientific insights and providing hints to detect potential new topics of interests. Moreover, an appropriate formal productivity model for evaluating infrastructure design decisions was developed. The result of the realization of this vision, a key contribution of this effort, is called the High-Productivity Framework that was tested and evaluated using real life-science application domain addressing breath gas analysis applied e.g. in the cancer treatment.

Keywords

productivity model; scientific analysis; scientific studies; cloud computing; security; breath gas analysis

Released

28.07.2016

Publisher

IEEE

Location

NEW YORK

ISBN

978-953-233-088-5

Book

2016 39th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO)

Pages from

186

Pages to

191

Pages count

6

URL

Documents

BibTex


@inproceedings{BUT163535,
  author="Peter {Brezany}",
  title="Towards a Novel Infrastructure for Conducting High Productive Cloud-Based Scientific Analytics",
  annote="The life-science and health care research environments offer an abundance of new opportunities for improvement of their efficiency and productivity using big data in collaborative research processes. A key component of this development is e-Science analytics, which is typically supported by Cloud computing nowadays. However, the state-of-the-art Cloud technology does not provide an appropriate support for high-productivity e-Science analytics. In this paper, we show how productivity of Cloud-based analytics systems can be increased by (a) supporting researchers with integrating multiple problem solving environments into the life cycle of data analysis, (b) parallel code execution on top of multiple cores or computing machines, (c) enabling safe inclusion of sensitive datasets into analytical processes through improved security mechanisms, (d) introducing scientific dataspace-a novel data management abstraction, and (e) automatic analysis services enabling a faster discovery of scientific insights and providing hints to detect potential new topics of interests. Moreover, an appropriate formal productivity model for evaluating infrastructure design decisions was developed. The result of the realization of this vision, a key contribution of this effort, is called the High-Productivity Framework that was tested and evaluated using real life-science application domain addressing breath gas analysis applied e.g. in the cancer treatment.",
  address="IEEE",
  booktitle="2016 39th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO)",
  chapter="163535",
  doi="10.1109/MIPRO.2016.7522135",
  howpublished="online",
  institution="IEEE",
  year="2016",
  month="july",
  pages="186--191",
  publisher="IEEE",
  type="conference paper"
}