Comments on: 5 trends that are changing how we do big data

By: Blake Oliver

Blake Oliver — Fri, 21 Dec 2012 12:09:02 +0000

It’s closer than you think!

http://bigdata.pervasive.com/

ENJOY!!

By: Blake Oliver

Blake Oliver — Fri, 21 Dec 2012 12:05:34 +0000

I think the most interesting thing in big data right now is using it for predictive analytics! The concept that you are able to use big data to get these predictions, from massive amounts of data divided into insane numbers of columns in a matter of minutes, is a priceless advantage for any business and Pervasive Software’s Rush Analytics is doing just that! And on top of that, they are giving away the $1000/month desktop version for FREE right now on their website and the trial last for a whole month! Check it out! You can download it right on to your system!! ENJOY!
http://bigdata.pervasive.com/Products/Download-Center.aspx

By: Phong Ngo

Phong Ngo — Thu, 15 Nov 2012 16:11:38 +0000

Right on: Big data without tool to help users understand and levarage in a meaningful way and, performance wise, reallistically will remain just that – big data. Users have the right to expect at least information if not knowledge out of big data. Analytics tools working closely with big data should become the tools of the masse, not just for a few so-called data scientists. That should be the next trend of Big Data. SQL-like on top of Hadoop/MapReduce in an on-line interactive is a good start.

Someday, with Big Data and popularized analytics tools, one should be able to ask: “What question should I ask” from this big data, instead of “how may times that name appears in this corpus of data.”

By: Lloyd B Hopkins

Lloyd B Hopkins — Fri, 09 Nov 2012 15:31:28 +0000

so is NetNow

By: Simon

Simon — Thu, 08 Nov 2012 16:23:09 +0000

Don’t forget Predixion Software. The team is the former data mining team from Microsoft and have developed a cloud architected big data machine learning platform. They can push their Excel generated models (MS or Mahout) directly into Hadoop, SQL, Greenplum etc and it’s much easier.

By: Mike

Mike — Tue, 06 Nov 2012 23:25:55 +0000

The following quote from an earlier post illustrates a key weakness of the “big data” movement.

“Their built-in analytics libraries for Machine Learning and BI integration provide a complete integrated solution from data ingestion and data processing to data delivery.”

What about data creation? Who is managing that? What’s needed is not more “analysts” who simply react to existing data. Instead, we need “researchers” who take an active role in determining what data are required to answer our critical business questions and how to best obtain that information.

Companies who use informed hypotheses to dictate what information is gathered are much less likely to find themselves drowning in “big data.”

By: Dev Bhatia

Dev Bhatia — Tue, 06 Nov 2012 15:05:03 +0000

Bringing features enabled by big data down to the hands of small businesses and consumers is absolutely the next step. At SRCH2 (http://srch2.com), we’re working on enabling small and mid-sized e-commerce retailers to offer high-end full-text search, and the many features that they are not yet tapping. These include: fuzzy search, full-text search, rapid geo-search, real-time updates, and much more. There’s still a whole lot left to do.

As for opportunists, that is clearly a risk. One of the things we find is that many stretch the definition of “big data,” and many also offer tools which are warmed over and modified versions of existing search software. These approaches are limited, as backwards integration of existing solutions leads to sub-optimal performance. If you have gone through the trouble of creating a big data stack, with several new elements built for speed and size, the last thing you want to do is have your search be the new bottleneck.

By: Dennis D. McDonald

Dennis D. McDonald — Tue, 06 Nov 2012 15:01:27 +0000

The angle I am researching is not “big data” per se but = how the data are generated in the first place, e.g., by government agencies that in the course of their legislatively mandated programs produce data that can be used by their target users as well as by others. More here: “A Framework for Transparency Program Planning and Assessment” http://www.ddmcd.com/outline.html

By: Jonathan

Jonathan — Tue, 06 Nov 2012 09:42:40 +0000

As mentioned in some of the discussion above I think a key issue is that BigData is no longer just the domain of the large companies. Companies of all sizes now facing large amounts and importantly a large variety of data many companies are not in a position to be hiring data scientists to help deal with their data problem.

At BIME moving forward we are very excited about BigData analytics in the cloud – Google BigQuery offers an analytical database as a service that scales to petabytes of data. It means companies that previously would have needed very large infrastructure and an operational team can now analyze their data with only a web browser. http://bigquery.bimeanalytics.com/

By: Melanie Jones

Melanie Jones — Tue, 06 Nov 2012 01:36:33 +0000

I agree that it’s important to know what the right questions are to ask. We have so much information at our finger tips, even without some of these new tools, it’s always a good idea to take a step back and think… “What decisions can I actually make with this knowledge? Is this important?”

And of course… the eternal favorite question of service providers and consultants (at least the good ones) – “Why?”