Can an SOA Approach Help Organizations Manage or Even Stay Ahead of the Impending 'Big Data' Wave?

"Data services" are of course just another SOA construct, albeit one that has been around since databases were invented. What might be more important is to deal with the data as it occurs, using event processing techniques, rather try and avoid drowning in the store-and-query-later approach that will stress data handling resources even more.

Not exactly. Data Services serve Business functions and cannot significantly affect the 'Big Data' wave.

Even for BI, where raw data is needed in large amount, Data Services do not dictate how large these data sets should be. Data Service deal with data semantics and ontologies rather than 'Big Data' waves.

Everything has to be done professionally and SOA is not necessary the ideal methodology for Data Management.

Well, a case in point... just look at Data.gov and this supporting article from David Linthicum last year at http://soa.sys-con.com/node/925170.

What do you think?

SOA, if combined with MDM, can improve data quality in operational systems.

Don't get confused with architecture and solving the business problem. There is no reason not to do new development that doesn't conform to well governed approaches to SOA. And, if you are planning a SOA/BPM project, you better plan on addressing data quality and governance as part of it, or you will not get the business results you are looking for. In fact, we are seeing that helping customers aligning data governance with business processes is one of the leading areas of opportunity for SAP with regards to SOA and Enterprise Information Management (EIM)

Big data -- flowing in at petabyte proportions -- requires a massively parallel infrastructure. Tools that are used for the most part now, such as spreadsheets, or desktop OLAP apps, will be woefully inadequate. What will be needed are tools/interfaces delivered virtually from the network, and available across the enterprise. These tools/interfaces need to be delivered as services, that can be adjusted and scale to handle large workloads.

Ultimately if you have well designed data services, it is the ultimate base upon which to build any application. In my experience, the big problems we see with any project be it 'Big Data' or a 'new app' is that the underlying data is a mess and people just continue to build on 'the mess'. If more of a focus was put on the very basic data services from the ground up, a lot of projects would run a lot more smoothly including concepts like 'big data'.

Big data are an issue of its own and is largely unrelated with SOA.
The "Big Data" is no news at all, and the technology required to manage it is already mature; it has been applied for decades to process data from high energy physics experiments at CERN or Lawrence Livermore etc.

Yes, but I believe it needs to be aligned with well-formed Data Governance Programs that can wrestle the data issues into submission so that well-formed logical models can be built and used by services. Of course, I may be just to consumed with data to make such an assumption :)