Dive into data

Take a closer look at how data an­a­lyt­ics is shap­ing the fu­ture of so­ci­ety.

The Star Malaysia - Star2 - - FRONT PAGE - Sto­ries by SU­SANNA KHOO bytz@thes­tar.com.my

IT’S time to set the record straight: Your per­sonal data re­ally isn’t as valu­able as you think it is. “Each of us, when we look at our per­sonal data, we think it’s worth a lot. But if I’m a (data) ag­gre­ga­tion com­pany, your per­sonal data is worth next to noth­ing,” says Brian Pren­tice, re­search vice pres­i­dent at Gart­ner Inc. “It’s only worth some­thing when it’s put to­gether with mil­lions of other people’s data.”

But that is not all. He points out that from 2014 on­wards many of us may soon find our­selves will­ing to vol­un­teer data about our­selves to com­pa­nies out there due to the in­cen­tives that we are of­fered in re­turn.

For in­stance, Pren­tice shares an ex­am­ple where shop­pers have the op­por­tu­nity to en­joy exclusive of­fers in var­i­ous stores within a par­tic­u­lar mall when­ever they are logged in to the free WiFi net­work pro­vided.

At the same time, by con­nect­ing to such a net­work, these shop­pers are agree­ing to al­low the oper­a­tors of the mall to col­lect per­sonal de­tails about them­selves and to track their move­ments around the mall.

De­spite an aware­ness of their data be­ing recorded by the mall oper­a­tors, he feels that most shop­pers would still choose to log onto the WiFi con­nec­tion, as they are able to ben­e­fit from spe­cial dis­counts and pro­mo­tions by do­ing so.

“By 2017, 80% of con­sumers will col­lect, track and barter their per­sonal data for cost sav­ings, con­ve­nience and cus­tomi­sa­tion,” Pren­tice says.

Ac­cord­ing to Gart­ner, three sec­tors that would be par­tic­u­larly af­fected by this trend would be on­line re­tail businesses, com­mu­ni­ca­tions ser­vice providers and fi­nan­cial ser­vices. It also said that com­pa­nies could shield them­selves from any po­ten­tial op­po­si­tion from con­sumers by first seek­ing ex­plicit con­sent from them be­fore pro­ceed­ing to col­lect their per­sonal data.

An­other likely sce­nario where it would be ben­e­fi­cial for you to share data about yourself to a third party would be in the case of pub­lic pol­i­cy­mak­ing. The United Na­tions’ (UN) Global Pulse ini­tia­tive is an ex­am­ple of this.

Es­tab­lished in 2009, it ex­plores how dig­i­tal data and real time an­a­lyt­ics tools can be used hand in hand to help pol­i­cy­mak­ers ef­fi­ciently track and deal with var­i­ous so­cio-eco­nomic is­sues such as poverty, hunger and the spread of dis­ease.

“While long term de­vel­op­ments (within a na­tion) still rely largely on trackers such as house­hold sur­veys and na­tional cen­sus data, au­thor­i­ties need to have ac­cess to re­li­able, real-time in­for­ma­tion on a big pic­ture level that would aid them in re­spond­ing quickly to any prob­lems that arise,” ex­plains I-Sah Hsieh, global man­ager of in­ter­na­tional de­vel­op­ment at an­a­lyt­ics soft­ware de­vel­op­ment com­pany, SAS In­sti­tute.

Through the anal­y­sis of pub­licly avail­able data such as con­ver­sa­tions con­ducted via so­cial me­dia, telecom­mu­ni­ca­tions call records and more, he says gov­ern­ments would be able to pre­dict up­com­ing eco­nomic trends. Con­se­quently, they would then be able to plan bet­ter in terms of de­vis­ing suit­able poli­cies to safe­guard a na­tion’s in­ter­ests de­spite un­avoid­able cir­cum­stances such as an im­pend­ing eco­nomic down­turn.

“In the past, this (anal­y­sis of big data) sim­ply wasn’t pos­si­ble due to the na­ture of con­ven­tional in­for­ma­tion collection and anal­y­sis meth­ods which are both la­bo­ri­ous and time con­sum­ing,” Hsieh says.

“Us­ing SAS So­cial So­cial Me­dia An­a­lyt­ics and SAS Text Miner... en­ables re­searchers to ac­cu­rately and ef­fi­ciently track all the chat­ter that was tak­ing place in the on­line space as it hap­pens.

Once cap­tured, re­searchers can then an­a­lyse the documents and ma­te­ri­als ob­tained to quan­tify feel­ings, moods, con­cerns and strate­gies. They can also use this data to cor­re­late with of­fi­cially avail­able sta­tis­tics to see how it matches up.”

De­sen­si­tised to data

Be­sides this, Gart­ner also ex­pects gov­ern­ments around the world to loosen up about what has pre­vi­ously been re­garded as clas­si­fied in­for­ma­tion in 2014 and be­yond.

By 2020, it fore­sees that en­ter­prises and gov­ern­ments will fail to pro­tect 75% of sen­si­tive data, caus­ing them to re­sort to de­clas­si­fy­ing the data and grant broad pub­lic ac­cess to it.

“There’s a huge ben­e­fit in open­ing up stuff that re­ally isn’t that se­cure in the first place,” says Pren­tice.

“So we see gov­ern­ments around the world look­ing at open data ini­tia­tives. And when you do that, people can start to use this in­for­ma­tion in ways that is use­ful to the govern­ment.”

For in­stance, if data was re­leased to the pub­lic re­gard­ing the kind of dis­eases that the govern­ment is re­search­ing on, he said that par­ties who had the rel­e­vant ex­per­tise could ac­tu­ally con­trib­ute to­wards these ef­forts.

The same prin­ci­ples on data clas­si­fi­ca­tion can be ap­plied to com­pa­nies as well, in Pren­tice’s opin­ion.

“The more we try to lock stuff down, the more it runs counter to driv­ing value back for the busi­ness,” he says.

Labour­ing the point

Be­sides data and pri­vacy is­sues, Pren­tice says that on­go­ing global dig­i­tal­i­sa­tion trends will re­sult in a “labour re­duc­tion ef­fect” which would, by the year 2020, “cause so­cial un­rest in a quest for new eco­nomic mod­els in sev­eral ma­ture economies”.

“The labour in­ten­sity in jobs will be di­luted by tech­nol­ogy,” he says. “It’s cre­at­ing new op­por­tu­ni­ties, but it is also de­stroy­ing a lot of jobs. I am hard pressed to find any in­dus­try right now that is not be­ing im­pacted by it. I think the only de­bate here is the level of dig­i­tal­i­sa­tion that’s go­ing on, not whether it’s hap­pen­ing or not.”

No­table eco­nomic sec­tors that would be im­pacted by this predica­ment in­clude the con­struc­tion in­dus­try, which re­lies heav­ily on man­ual labour.

How­ever, knowl­edge work­ers are in no ways ex­empted from this wave of change ei­ther.

“By 2020, the ma­jor­ity of knowl­edge worker ca­reer paths will be dis­rupted by smart ma­chines in both pos­i­tive and neg­a­tive ways,” says Pren­tice.

The im­pact to the knowl­edge worker is that if you’re the type whose job is to pro­vide an­swers to people or your value is be­cause you have cer­tain in­for­ma­tion or knowl­edge that oth­ers don’t, then your value be­comes marginalised.”

This is be­cause of the an­tic­i­pated in­crease in de­pen­dence on ar­ti­fi­cial in­tel­li­gence sys­tems such as IBM’s Wat­son which uses nat­u­ral lan­guage pro­cess­ing and an­a­lyt­ics to process in­for­ma­tion and help in de­ci­sion mak­ing.

“Credit Agri­cole pre­dicts that Wat­son de­rived sys­tems will ac­count for 12% of IBM’s to­tal rev­enue by 2018,” Pren­tice says.

Since its de­but on tele­vi­sion in 2011, Wat­son’s sys­tem per­for­mance is 24 times faster and its size is now 90% smaller.

The cog­ni­tive com­puter sys­tem can now be op­er­ated on a sin­gle IBM Power 750 server us­ing Linux, thus tran­si­tion­ing from its orig­i­nal size which was that of a mas­ter bed­room to some­thing much smaller — the equiv­a­lent of four pizza boxes. In ad­di­tion, its ser­vices are now de­liv­er­able via cloud of­fer­ings and on­line chat ses­sions.

Wat­son is now be­ing de­ployed across a wide range of in­dus­tries in­clud­ing health­care, bank­ing and telecom­mu­ni­ca­tions via the com­pany’s Cus­tomer Early En­gage­ment Pro­gram, which was launched in the first quar­ter of 2013.

Or­gan­i­sa­tions across the globe in­clud­ing lo­cal telecom­mu­ni­ca­tions com­pany, Cel­com Ax­i­ata, Sin­ga­pore-based DBS Bank, ANZ Bank­ing Group, IHS, Nielsen and the Royal Bank of Canada are among those par­tic­i­pat­ing in pi­lot tests us­ing Wat­son’s tech­nol­ogy.

Through these tri­als, Wat­son’s abil­ity to un­der­stand the nu­ances of hu­man lan­guage, im­prove its per­for­mance through con­tin­u­ously learn­ing based on a user’s be­hav­iour, process com­plex ques­tions, and search vast amounts of big data to pro­duce rel­e­vant, ev­i­dence based re­sponses will be fine tuned to en­able com­pa­nies to bet­ter un­der­stand and cater to their cus­tomers’ needs.

In a re­port en­ti­tled Gart­ner top pre­dic­tions 2014: Plan for a dis­rup­tive, but con­struc­tive fu­ture, Gart­ner pre­dicts that Wat­son will ac­count for at least 1.5% of IBM’s rev­enue by the end of 2015, and would then in­crease to 10% by the end of 2018.

Dig­i­tal shift: de­mand for labour will be­come di­luted as it gets re­placed by tech­nol­ogy.

Pub­licly avail­able data can help gov­ern­ments bet­ter deal with so­cioe­co­nomic is­sues. IbM Wat­son so­lu­tions vi­cepres­i­dent Stephen Gold us­ing Wat­son, which helps businesses and cus­tomers with their de­ci­sion mak­ing.

Per­sonal data be­comes use­ful to businesses only when it is an­a­lysed in the mil­lions.

Con­sumers may be will­ing to share per­sonal data when of­fered in­cen­tives in re­turn, such as exclusive dis­counts and of­fers.

Newspapers in English

Newspapers from Malaysia

© PressReader. All rights reserved.