Objectives

 

The Big Data Analytics (BDA) Interest Group seeks to develop community based recommendations on feasible data analytics approaches to address scientific community needs of utilizing large quantities of data. The objectives are complementary to the recent activities performed by ‘NIST big data’ working groups and supports formation of different working groups to tackle specifics problems. The overall objectives of the group are:

 

• BDA will aim to clarify some foundational terminologies in the context of data analytics understanding differences/overlaps with terms like data analysis, data mining, etc.

• BDA will develop a recommendation documents with a systematic classification of feasible combinations of analysis algorithms, analytical tools, data and resource characteristics and scientific queries. These recommendation documents can serve as a best practice guide for scientific groups/communities interested in investing in Big Data technologies

• BDA will work towards a consensus amongst its members to achieve this desired goal  

• BDA will collaborate with external bodies and initiatives – such as NIST, OGC, ISO, EarthServer and others – for exchange; to this end, BDA findings will be made available for download publicly and freely. Specifically, our interaction with NIST Big Data working groups will entail:

  • BDA will encourage discussions linking NIST activities and reference architecture-related terminologies. BDA will focus on analyzing selected use cases (with overlap with NIST [1]) targeting data analytics/analysis approaches (e.g. map-reduce, array databases, computational statistics, etc.)

  • BDA will support formation of working groups from NIST and BDA to work together on subtopics.

 

Participation

BDA is open to all RDA members to participate. BDA welcomes individuals with the following expertise to actively participate in its activities

• Domain science experts grappling with Big Data as part of their scientific applications

• Database, HPC and distributed computing experts with middleware experience such as Hadoop and NoSQL/NewSQL

• Analytics experts with tools/algorithmic expertise, including data mining and machine learning approaches

• Managers running or seeking to deploy Big Data solutions at their organizations

• Members of groups that participate in other, more broad activities, e.g. like the US NIST big data working groups

 

Outcomes

BDA will be considered a success if the Interest Group:

• Develops recommendation documents based on consensus amongst its members and in collaboration with other globally related activities (e.g. NIST, or scientific community-specific associations and organizations like EarthServer)

• Visibility and uptake of results can be demonstrated within RDA, but also beyond RDA (such as OGC and ISO, for example)

 

Mechanism

BDA will utilize capabilities provided by the RDA platform to effectively communicate and collaborate. These include:

• Monthly telecoms/webex to with planned agenda to discuss specific issues

• Asynchronous collaboration using google docs, wiki and email list servers

• RDA meetings to hold sessions for F2F interactions amongst members and to inform other RDA members of its ongoing activities.

All BDA documents will be publicly available.

 

Timeline

A tentative timeline is given below, but is subject to change.

 

Until 3rd RDA Plenary (03/2014):

• Terminology definition

• Science Use Cases

• Use Case Template

• Mapping Science Use Cases to the Template

• Technology Lists

Until 4th RDA Plenary (tbd):

• Technology Template

• Mapping Technologies to the template

• Mapping Uses cases to Technology Solutions

• Recommendation document

 

List of Participants

• Rahul Ramachandran (chair)

• Morris Riedel (chair)

• Peter Baumann (chair)

• Ahmed Shiraz Memon

• Andre Schaaff

• Bohdan Schatschneider

• Byron Georgantopoulos

• Cynthia Hudson Vitale

• Francine Berman

• Giuseppe Andronico

• Giuseppe Fiameni

• Hilary Hanahoe

• Hugh Shanahan

• Ian Truslove

• Leif Laaksonen

• Luca Di Fiore

• May Yuan

• Michael Bruck

• Robert Marcus

• Stephen wolff

• Timea Biro

• Wo Chang

• Yannis Ioannidis

• Yde de Jong

 

References

[1] NIST Big Data Use Cases

http://bigdatawg.nist.gov/usecases.php