Month: November 2009

Video on Big Data

A terrific video on challenges of handling Big Data by ComScore

comScore is the global source of digital market intelligence and the most preferred measurement service. Clients turn to comScore for objective, accurate and reliable insights into consumers’ online behavior and for details into their demographic characteristics, attitudes, lifestyles and offline activities.

http://www.asterdata.com/customers/comscore-video.php

It is a good educational video (simple and short) for people wanting to understand what exactly does SQL and MapReduce integration helps in analytics. Basically with bigger data, ease of querying is as important if not more as most descriptive statistics consists of summarization and slicing and dicing appropriate cuts of data. When it comes to predictive analytics or hypothesis testing the traditional statistics requires tweaking of their assumptions including continuity corrections and confidence intervals can be trickier if not handled properly.

I met the chaps at Com Score at Big Data Summit and they seemed to have a tremendous amount of data- literally measuring the whole internet everyday.

Student Statement: The Right to Research

An initiative by Student Government at U Tenn ( I am a slight pat of Student govt but not of the following ———as  a member of the University of Tennessee Technology, Fee Advisory Board. My current role involves increasing funding for bears like koala 😉  )

Scholarly knowledge is part of the common wealth of humanity.

Unfortunately, not everyone has access to the scholarly literature, despite advances in communications technology.  The high cost of academic journals restricts access to knowledge; in some fields, prices can reach $20,000 for a single journal subscription1 or $30 for an individual article.2 Despite these high prices, authors of scholarly articles are not paid for their work. The profits from these publications go solely to the publishers of the journals.  A vast amount of research is funded from public sources – yet taxpayers are locked out by the cost of access.

Screenshot

I suppose companies like SAS Institute ( with a nice SAS Publishing arm- I got a SAS Enterprise Guide book for predictive analytics from them) Aster Data ( which needs all the BIG DATA programmers and researchers including students), SPSS ( with IBM’s backing and pedigree of R and D) , and SAP (with University Network) and even the dropout* founded Oracle

can help by sponsoring journal articles so as to

1) Increase pool of developers who remain loyal to that platform for life ( similar to companies offering student credit cards)

2) Increase visibility as a low cost advertising medium.

 

( *Amazing- Google, Microsoft, Oracle, Trilogy, Aster (partly) , JMP (partly) it seems to get really really rich- one has to go to Grad School, make a tech company and drop out.

Maybe I do a research paper on this hypothesis using some kind of ANOVA, T tests)

If you believe students have Right to Research and you can help by stepping in to help both article authors and students come closer AND makes good sense for your business

– HONK YOUR HORN.

A website for God

Here is a website I built in less than 15 minutes for my fellow Bible Study and Church members. I am hoping to introduce the concepts of search engine optimization and social media to the people who have prayed for me all this year in 2009 and helped me out of very torrid personal times and battles.

It is called

Prayers2Go.com

Basically: you order dinner or burgers for friends all the time. Maybve all they ever needed was a prayer.

For Here or For to Go.

Oracle Open World and Techie Events

An innovate way to showcase collateral thanks to Oracle Open World.

The Post Event has an easily searchable archive with downloadable files and partner collateral.

See this-

http://myexpospace.com/oracle2009/vcr2009/vcr.cfm?search=advance

untitled

An even better site is their streaming site post event-( which is better than an HTML website —isnt it)

http://ondemandpreview.vportal.net/

oracle

However the website-http://myexpospace.com/ which has enabled Oracle to do this Post Event Content Management seems to be in a closed beta (as they update their websites)

Post event content management helps in getting out the word to people who were unable to see it, and the analytics from site visitor behavior can help you gauge viewer interest. I personally also like the concept of Virtual Conferences  as well – as they can be done much more frequently than they are currently are-

Like the Cloud Slam event I was invited to speak earlier ( and missed because of time zone differences !)

https://decisionstats.wordpress.com/2009/04/10/cloud-nine/

and https://decisionstats.wordpress.com/2009/03/12/here-comes-the-cloud-slam/

DecisionStats :Lies , True Lies and Statistics

In keeping with a promise-

here are transparent Decisionstats.com stats-

I now have 5898 views in October (

I moved from a self hosted WordPress site to a WordPress.com site that I regularly backup the export file- because of repeated hacking attempts)

Thanks to bad health- I can no longer travel to conferences- I recommend that you do a live streaming of the videos and all sessions. Saves the environment lots of carbon debt. Expands your potential audience by a factor of 10. Saves you money, too.

Ads- Its almost Christmas- so here comes my Hat- 200$ per month for ads, $25 for newsletter ads ( goes to 450 people PER post) . 3months minimum.

Not accepting exclusive ads to maintain editorial integrity and neutral behavior. Not accepting Click per Conversion Ads ( Does this look like Britney Spears blog?) This means if I shout on the phone to certain Mr Smith , I can do so with a clear conscience ( and voice).

Money goes to supplement my research. I am currently working on HPC and parallel algorithms ( once I get off sick leave). I am trying out parallel R on MPP clusters ( but also studying my latest acquisition a book by SAS Enterprise Miner ( courtesy the lovely elegant SAS Publishing people )  and if it can be done on an Open MPI framework ( probably can’t).

Q- If God was a Blogger, What would HE say?

 

DecisionStats

 

A- God wouldnt say anything if HE was a blogger.

He would write it all down instead.

 

 

Tera Data and SAS innovate together

I missed out posting this one- it’s the big big news of Tera Data and SAS coming closer to create a much needed Business Analytics Centre.

SAS and Teradata will establish a Business Innovation Analytic Center.

In conjunction with Elder Research Inc., the Center will offer leading-edge analytic thinking and implementation.

Core news facts:

With this centralized “think-tank,” customers can discuss analytic best practices with domain subject matter experts and quickly test or implement innovative models focused on uncovering unique insights for optimizing business operations.

The Business Analytic Innovation Center will combine the strengths of SAS, Teradata and Elder Research (the world’s leading analytical consulting firm for data mining and predictive analytics) to help customers across all industries grow revenue, reduce risk and improve operations.

The Center will incorporate unmatched thought leadership with a visionary lab for pilot programs, analytic workshops, and proofs-of-concept for prospective customers across a myriad of industries, including financial services, retail, government, health care and life sciences, and insurance.

Also see- http://www.sas.com/partners/directory/teradata/index.html

teradata1

Press Release

CARY, NC (Oct. 26, 2009) – Recognizing the growing need and challenges businesses face driving operational analytics across enterprises, SAS and Teradata are planning to establish a centralized “think tank” where customers can discuss analytic best practices with domain and subject-matter experts, and quickly test or implement innovative models that uncover unique insights for optimizing business operations. The Business Analytics Innovation Center will combine the strengths of SAS, the leader in business analytics <http://www.sas.com/businessanalytics/> software and services, Teradata Corporation <http://www.teradata.com> (NYSE: TDC <http://www.nyse.com/about/listed/lcddata.html?ticker=TDC> ), the world’s largest company solely focused on data warehousing <http://www.teradata.com/t/enterprise-data-warehousing/> and enterprise analytics <http://www.teradata.com/t/business-needs/data-mining-and-analytics/> , and Elder Research Inc <http://www.datamininglab.com/> . (ERI), the world’s leading analytical consulting firm for data mining and predictive analytics, to help customers across all industries grow revenue, reduce risk and improve operations.

“For decades, SAS, Teradata and ERI have provided state-of-the-art analytical solutions for their respective customers,” said Jim Davis, SAS Senior Vice President and Chief Marketing Officer. “The Business Analytics Innovation Center combines our individual strengths for a robust approach to enterprise analytics. By acting faster and asking questions they previously couldn’t ask or didn’t think of, participating clients will be able to improve their operations and strengthen competitive advantage.”

The Center will incorporate unmatched thought leadership with a visionary lab for pilot programs, analytic workshops, and proofs of concept for prospective customers across a myriad of industries, including financial services, retail, government, health care and life sciences, and insurance.

The joint offerings that will comprise the Center are expected to yield powerful results. For example, a top global consumer electronics firm used resources from ERI that will become part of the Business Analytics Innovation Center to discover and prevent fraud committed by clients and partners. Within the first year, the firm recovered more than $20 million.

They continue to rely on ERI for new analytical programs, rewarding their anti-fraud division’s effectiveness with expanded budget, staff and charter.

(Note from Ajay- Read earlier article on case management solutions by SAS in recent press release announcement at Business Leadership series ( the event following the Data Mining 2009 series)

Screenshot- Terrific web and social media campaign by www.teradata.com
( The Only BI company I know of that has an Iphone application AND a Facebook application)

Aster Data : Big Data Bigger Analytics Campaign

My favorite ( as of now) company in Big Data is Aster Data* ( I am partial to companies founded by Stanford Alumni having interacted with a lot of them while working with Trilogy- another Stanford dropout alumni company. There are also not too many Silicon Valley startups by us famously non intellectual Punjus

Q- What is the culture in Punjab A- In Punjab the only culture is agriculture)

Aster Data has correctly hit the marketing hammer on the nail of bigger data and with the quantities of data expanding rapidly  this is a lucrative market to get into ( as pointed by our favorite analytics journal NY Times)

Aster Data’s products of nCluster and nPath with MapReduce SQL, and the recent interactions with SAS Institute hold them in a nice promising place but with miles to go before they even rest ( or start thinking of that IPO)

Aster were present at Data Mining 2009 with terrific response to their booth.

As a techie wannabe stats frat boy, I like the Aster nPath product more (Time Series)  but the analytics within database claim with nCluster needs to be investigated and even tested further. Especially if you need  three days to get your monthly summary report.asterdata

( *and also an advertiser, sponsor to Big Data Summit as per FCC regulations)

The Data Services and Applications with flexibility for cloud computing is what makes this especially appealing from a product perspective while their relatively small size ( as compared to other bigger Vend- ORs) gives alliance partners more leverage in colloborating in Research and Design and maybe even co bundling applications.

Screenshots- Courtesy -The Lovely www.asterdata.com website ( Webmasters of other websites especially IBM and Oracle’s should take note how a website can have lots of content and yet be readable)

Also I will be posting the remaining Data Mining 2009 interviews shortly (including Part 2 with Anne) and share some/all of the presentations  via SlideShare embedding in WordPress.com ( post permission).

As for the Aster Data Interviews- I owe Peter Pavloski and the readers one. Coming up soon.

asterdata2