Data science news

How India’s fa­vor­ite TV show uses data to change the world

Satyamev Jayate, one of India’s highest-rated tele­vi­sion shows, is us­ing data as a means to ef­fect mean­ing­ful change. The show’s pro­du­cers are ag­greg­at­ing and ana­lyz­ing the mil­lions of mes­sages they re­ceive on con­tro­ver­sial is­sues to do everything from plan­ning fu­ture epis­odes to push­ing for polit­ic­al change.

Data Science in India: meet­ing re­quire­ments, not just budgets

By Andrew Brust (Contributor at ZDNet)

It struck me that edu­ca­tion­al val­ues and ap­proaches in India might make Data Science skill sets there more abund­ant than in the U.S. and oth­er de­veloped coun­tries. Perhaps Data Science will help India tran­scend the stigma/typecast of a tech tal­ent center that is merely lower in cost.

India’s edu­ca­tion cul­ture is be­ne­fit­ing Data Science, and the coun­try is pro­du­cing Data Analytics pro­fes­sion­als that com­pete not just on re­l­at­ive cost, but on ab­so­lute tal­ent.

From the hub of the IT ser­vices sec­tor, the shift is now to­ward mak­ing India an en­gine for knowledge-based ser­vices, with Data Analytics fore­most among them.

Data de­coded 

Data, es­pe­cially Big Data, is key to de­cision mak­ing and com­pan­ies world­wide are start­ing to value this in their busi­ness. However, the mere volume of data is not the is­sue that or­gan­iz­a­tions face today. It is the un­struc­tured nature of the data and the chal­lenge in ex­tract­ing busi­ness value from it that is bey­ond the reach of tra­di­tion­al en­ter­prise tools and busi­ness prac­tices. According to a McKinsey study , the US alone faces a de­fi­cit of 140,000-190,000 data sci­ent­ists again­st a pro­jec­ted de­mand of 440,000-490,000 by 2018. A short­age of the ana­lyt­ic­al and ma­na­geri­al tal­ent ne­ces­sary to make the most of Big Data is a sig­ni­fic­ant and press­ing chal­lenge and one that com­pan­ies need to ad­dress. Keeping this in mind, many Fortune 500 firms have already star­ted to lever­age India and its Data Science pro­fes­sion­als to com­pet­it­ive ad­vant­age.

Destination, not de­tour 

While many US firms were ini­tially hes­it­ant to work with an off­shore pro­vider, sourcing to India is no longer a ta­boo. But some are still skep­tic­al about shift­ing work geo­graph­ic­ally for lower costs. For Data Science, India is not simply a cheap­er al­tern­at­ive; it’s a go-to mar­ket for tal­ent that can’t be found else­where.

Data Science and Big Data may mark a turn­ing point for India and, most likely, oth­er coun­tries where math­em­at­ics edu­ca­tion is heav­ily em­phas­ized. Ultimately, mar­kets that stress edu­ca­tion in ad­di­tion to tech­no­logy in­nov­a­tion will be well- po­si­tioned. India’s prov­ing that today.

How Big Data Startups Could Kill A $30 Billion Industry 

When it comes to big data, “size doesn’t mat­ter,” Ravi Mhatre, man­aging dir­ect­or of ven­ture firm LightspeedVenture Partners just told Business Insider.

“It’s not just big data. It’s got to be fast data and it’s got to be mean­ing­ful data” he says.

There’s a new wave of start ups try­ing to make it easi­er to use big data sys­tems. If they suc­ceed, they will really hurt the multi-billion dol­lar mar­ket for business- in­tel­li­gence soft­ware and threaten products like SAP’s Business Objects, IBMCognos and Oracle Hyperion.

“We’re talk­ing an in­dustry today that’s prob­ably $20 [bil­lion] to $30 bil­lion that I think, overnight, is go­ing to be re­placed by a com­pletely new set of plat­forms,” he pre­dicts.

Future of data ana­lyt­ics is pre­dict­ive, ac­tion­able

Move over, ret­ro­spect­ive data ana­lys­is – the fu­ture is in real-time and pre­dict­ive ana­lyt­ics, says a new mar­ket re­port from Frost & Sullivan. The trend is also to­ward web-based sys­tems that ag­greg­ate dis­par­ate data across di­verse care set­tings.

The more “hol­ist­ic” ap­proach to data min­ing in­cludes clin­ic­al data from elec­tron­ic health re­cords com­bined with fin­an­cial and ad­min­is­trat­ive in­form­a­tion to provide a more well-rounded view of the qual­ity and ef­fi­ciency of pa­tient care and then us­ing that in­form­a­tion to make stra­tegic de­cisions, ac­cord­ing to a Frost & Sullivan an­nounce­ment.

The firm pre­dicts that the use of ad­vanced health data ana­lyt­ics solu­tions in hos­pit­als will grow sig­ni­fic­antly to 50 per­cent ad­op­tion in 2016 up from about 10 per­cent last year. That’s a 37.9 per­cent com­pound an­nu­al growth rate and an in­crease of 400 per­cent over the baseline.

“Hospitals will in­creas­ingly in­vest in ad­vanced data ana­lyt­ics solu­tions to mon­it­or end-to-end care de­liv­ery across a vari­ety of set­tings,” Frost & Sullivan ana­lyst Nancy Fabozzi said. “Due to grow­ing com­pet­it­ive pres­sures, hos­pit­als need to provide com­pre­hens­ive re­port­ing on per­form­ance and qual­ity meas­ures to a vari­ety of stake­hold­ers. Advanced ana­lyt­ics cap­ab­il­it­ies are ab­so­lutely crit­ic­al for sur­viv­al there is no way to avoid it.”

Customer Analytics: Social and Predictive Gains a Few Years Off

In a Forrester sur­vey of 90 en­ter­prises cus­tom­er ana­lyt­ics prac­ti­tion­ers, in­clud­ing in­surers, so­cial ana­lyt­ics is the most pop­ular ana­lyt­ics pro­gram com­pan­ies are look­ing to pi­lot in the next couple years.

The Forrester re­port, “The State of Customer Analytics 2012,” in­volved an ex­tens­ive sur­vey of 90 en­ter­prise cus­tom­er ana­lyt­ics prac­ti­tion­ers in bank­ing, fin­ance, util­it­ies and pro­fes­sion­al ser­vices. In it, the ma­jor­ity of cus­tom­er ana­lyt­ics prac­ti­tion­ers in the re­port noted ma­ture use of re­port­ing and BI (69 per­cent), de­script­ive ana­lyt­ics (81 per­cent) and pre­dict­ive mod­els (73 per­cent) for cus­tom­er met­rics. That level of ad­op­tion and use puts ana­lyt­ic prowess in mar­ket­ing and sales above most oth­er en­ter­prise de­part­ments. Now, these de­part­ments plan to take their next for­ays in­to re­turns and de­velop their cus­tom­er data re­sources, with a some­what mixed bag of plans, says Srividya Sridharan, a cus­tom­er in­tel­li­gence pro­fes­sion­als ana­lyst at Forrester and an au­thor on the re­port.

When asked about the top cus­tom­er ana­lyt­ics pro­gram they planned to pi­lot in the next two years, “so­cial ana­lyt­ics” led the way, though 30 per­cent of re­spond­ents pegged so­cial ana­lyt­ics re­turns as a long-term goal. This is not be sur­pris­ing in terms of so­cial me­dia data in­terest, but def­in­itely “in­dic­ates that so­cial data is still a largely un­ex­plored data source” at the present, says Sridharan.

Sridharan says she ex­pects busi­nesses that have already made in­vest­ments in bet­ter cus­tom­er data man­age­ment and meas­ure­ments to take a ser­i­ous look at pre­dict­ive ana­lyt­ics as a next step. However, the Forrester ex­pert cau­tioned that en­ter­prises need to eval­u­ate what re­turns they would ex­pect from a pre­dict­ive plat­form be­fore diving in to an im­ple­ment­a­tion, as well as take an in­ward look at the in­form­a­tion at hand and how it would be man­aged mov­ing for­ward.

“One of the pre­requis­ites for pre­dict­ive ana­lyt­ics is to have the right type of customer-level data avail­able and ac­cess­ible at the ap­pro­pri­ate gran­u­lar­ity in or­der to build pre­dict­ive mod­els. While firms can pi­lot or build mod­els based on pre­dict­ive ana­lyt­ics tech­niques, where they need to fo­cus more is in put­ting these mod­els and scores to work through mar­ket­ing ex­e­cu­tion sys­tems that ac­tu­ally man­age cus­tom­er in­ter­ac­tions,” Sridharan says.

How CIOs Can Extract Value from Big Data

In a large re­tail chain, Big Data was used to tweak pro­duct pri­cing at the store level in real time, based on an evolving set of cri­ter­ia ran­ging from point-of-sale data, to email-based pro­mo­tions, to loc­al ad­vert­ising. Much like a doc­tor start­ing a dia­gnos­is by tak­ing tem­per­at­ure and blood pres­sure, then gradu­ally re­fin­ing each sub­sequent test, the re­port­ing gen­er­ated by Big Data rap­idly evolved as the “dia­gnos­is” pro­gressed, with mar­ket­ing driv­ing the ef­fort and em­bed­ded IT staff ad­just­ing the tech­nic­al side of the house in real time, based on the re­com­mend­a­tions of cross-functional ana­lyt­ic­al ex­perts. In short, mar­ket­ing asked the ques­tion, the ana­lysts de­term­ined what data were needed to an­swer it, and the IT people im­ple­men­ted the tech­nic­al as­pect.

Facebook Tackles (Really) Big Data With ‘Project Prism’

Facebook is star­ing down at lar­ger ava­lanche of data, and there are new lim­it­a­tions that need fix­ing. This week, dur­ing a brief­ing with re­port­ers at Facebook’s Menlo Park headquar­ters, Parikh re­vealed that the com­pany has de­veloped two new soft­ware plat­forms that will see Hadoop scale even fur­ther. And Facebook in­tends to open source them both.

The first is called Corona, and it lets you run myri­ad tasks across a vast col­lec­tion of Hadoop servers without run­ning the risk of crash­ing the en­tire cluster. But the second is more in­triguing. It’s called Prism, and it’s a way of run­ning a Hadoop cluster so large that it spans mul­tiple data cen­ters across the globe.

“It lets us move data around, wherever we want,” Parikh says. “Prineville, Oregon.Forest City, North Carolina or Sweden.”

Leave a Reply