Big Data, Big Impact:New Possibilities for International Development                                             0
Executive Summary  A flood of data is created every day by the interactions of billions of  people using computers, GPS de...
Big Data, Big Impact:New Possibilities for International DevelopmentBy analysing patterns from mobile phoneusage, a team o...
Likewise, utilising the data created by mobileData through the Mobile Financial Services                  phone use can im...
individuals voluntarily or through crowdsourcing.             for harnessing data for policy and action. ItsNGOs like Usha...
changing their behaviour and prioritise where             firms, governments, and individuals can beyou put resources in r...
sector firms do not see an incentive to share data     In the area of individual incentives, Jana, a Boston-   they regard...
framework, including protections for data reuse,                       providing the data online, governments need towere ...
available to better serve individuals in emergingmarkets should outweigh these risks.                                    ...
For more information please visit: World Economic Forum Mobile Financial Services Development Report www.weforum.orgmfs Gl...
Upcoming SlideShare
Loading in...5
×

Wef big databigimpact_briefing_2012

161

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
161
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
5
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Wef big databigimpact_briefing_2012

  1. 1. Big Data, Big Impact:New Possibilities for International Development 0
  2. 2. Executive Summary A flood of data is created every day by the interactions of billions of people using computers, GPS devices, cell phones, and medical devices. Many of these interactions occur through the use of mobile devices being used by people in the developing world, people whose needs and habits have been poorly understood until now. Researchers and policymakers are beginning to realise the potential for channelling these torrents of data into actionable information that can be used to identify needs, provide services, and predict and prevent crises for the benefit of low-income populations. Concerted action is needed by governments, development organisations, and companies to ensure that this data helps the individuals and communities who create it.Special ThanksThe World Economic Forum acknowledges the work of Vital WaveConsulting in assembling this briefing and the extensive researchconducted by partners such as the Boston Consulting Group, the UNGlobal Pulse, the Global Viral Forecasting Initiative, Ushahidi, theWorld Bank, Vodafone, and the World Economic Forum‟s ICT GlobalAgenda Council on the transformative potential for harnessing big data. 1
  3. 3. Big Data, Big Impact:New Possibilities for International DevelopmentBy analysing patterns from mobile phoneusage, a team of researchers in San Francisco Financial Servicesis able to predict the magnitude of a diseaseoutbreak half way around the world. Similarly, Data gleaned from mobile money services can provide deep insight into spending and saving habits acrossan aid agency sees early warning signs of a sectors and regions. Digital payment histories can allowdrought condition in a remote Sub-Saharan individuals to build credit histories, making themregion, allowing the agency to get a head start candidates for loans and other credit-based financialon mobilising its resources and save many services.more lives. EducationMuch attention is paid to the vital services that Data derived from the use of mobile value-addedmobile phone technology has brought to billions of services can be used to improve public-sectorpeople in the developing world. But now many understanding of educational needs and knowledgepolicy-makers, corporate leaders and development gaps, allowing more targeted and timely initiatives toexperts are realising the potential applications, like disseminate critical information.the examples above, for the enormous amounts ofdata created by and about the individuals who use Healththese services. Data collected through mobile devices, whether captured by health workers, submitted by individuals, or analysedSources such as online or mobile financial in the form of data exhaust, can be a crucial tool intransactions, social media traffic, and GPS understanding population health trends or stoppingcoordinates now generate over 2.5 quintillion bytes outbreaks (see box on page 5). When collected in the i context of individual electronic health records, this dataof so-called „big data‟ every day . And the growth of not only improves continuity of care for the individual, butmobile data traffic from subscribers in emerging it can be used to create massive datasets with whichmarkets is expected to exceed 100% annually treatments and outcomes can be compared in an iithrough 2015 . efficient and cost effective manner.The data emanating from mobile phones holds Agricultureparticular promise, in part because for many low- Mobile payments for agricultural products, inputincome people it is their only form of interactive purchases and subsidies may help governments bettertechnology, but it is also easier to link mobile- predict food production trends and incentives. Thisgenerated data to individuals. This data can paint a knowledge can be used to ensure the availability ofpicture about the needs and behaviour of individual proper crop storage, reduce waste and spoilage, andusers rather than simply the population as a whole. provide better information about what types of financial services are needed by farmers. Mobile use patterns may also help governments and developmentBuilding user-centric solutions offers compelling organisations identify regions in distress so that targetedpossibilities for providing better access to services in assistance can be directed to them. Early detection canhealth, education, financial services, and agriculture help prevent families from leaving their land and furtherfor people living in poverty. decreasing agricultural production. 2
  4. 4. Likewise, utilising the data created by mobileData through the Mobile Financial Services phone use can improve our understanding ofLens vulnerable populations, and can quicken governments‟ response to the emergence ofA look at the „pillars‟ from The World Economic new trends. Actors in the public, private, andForum‟s Mobile Financial Services Development development sectors are beginning to recogniseReport offers insights into how the requirements of the mutual benefits of creating and maintainingmobile financial services development coincide with a „data commons‟ in which this informationprerequisites for a thriving data commons. benefits society as a whole while protecting individual security and privacy. But a moreRegulatory Proportionality: Finding appropriate uses concerted effort is required to make this vision afor mobile-generated data will require regulation similar reality.to that needed for mobile financial services. In bothsituations, regulation must keep pace with new Understanding the Dynamics of thetechnology and protect consumers without stiflinginnovation or deterring uptake. The development of Data Ecosystemsensible data standards could increase uptake of both To turn mobile-generated data into an economicmobile financial services and individual data security. development tool, a number of ecosystem elements must be in place. For thoseConsumer Protection: As with mobile financial individuals who generate the data, mechanismsservices, proper regulation and data ownership must be developed to ensure adequate userprocesses must be put in place to prevent the theft or privacy and security. At the same time,misuse of sensitive information. business models must be created to provide theMarket Competitiveness: In the long term, adequate appropriate incentives for private-sector actorscompetition is essential to ensure a wider range of to share and use data for the benefit of society.affordable services and interoperability. However, Such models already exist in the Internetprivate-sector companies should be encouraged to environment. Companies in search and socialallow access to non-sensitive data that can benefit networking profit from products they offer at nopopulations and deepen their own understanding of charge to end users because the usage dataindividual behaviour. Such cooperation may also help these products generate is valuable to othertelecom operators realise that creating interoperable ecosystem actors. Similar models could bemobile money systems can benefit them over the long created in the mobile data sphere, and the dataterm. generated through them could maximise the impact of scarce public sector resources byMarket Catalysts: For both the data commons and indicating where resources are most needed.mobile money, government can serve as a catalyst toensure legitimacy. This will require open and A look at the various types of data and actors intransparent governance, as the idea of government the data ecosystem illustrates the roles andaccess to an individual‟s financial information could incentives at work. The private sector maintainsdiscourage uptake of mobile financial services. vast troves of transactional data, much of which is „data exhaust‟, or data created as a by-End User Empowerment & Access: Individuals must product of other transactions. With the use ofhave a moderate degree of financial literacy, affordable mobile phones, much of this data can beaccess to a mobile device, and a mobile network associated with individuals and their locations.connection, in addition to control over their own The public sector in most countries alsoinformation. maintains enormous datasets in the form of census data, health indicators, and tax andDistribution and Agent Network: Analysing expenditure information.transactional data could determine where there is The Internet and mobile revolution have addeddemand for additional mobile money agents. yet another source: data contributed byAdoption & Availability: Open data can helpdetermine which finance products are in the highestdemand, matching demand with supply. 3
  5. 5. individuals voluntarily or through crowdsourcing. for harnessing data for policy and action. ItsNGOs like Ushahidi are already using crowdsourcing director, Robert Kirkpatrick, says that datato obtain, verify and disseminate real-time collected through mobile device usage can spurinformation about natural disasters and election effective action in two primary ways: bymonitoring, and Ushahidi is developing ways to filter reducing the time lag between the start of aand use the huge amounts of information being trend and when governments and other iiicreated with applications such as SwiftRiver . With authorities are able to respond to them, and bythe overwhelming majority of the world now having reducing the knowledge gap about how people ivaccess to a mobile phone , crowdsourcing allows respond to these trends.individuals to contribute to the information gathering vprocess, making it more democratic and transparent. Kirkpatrick cites Dr. Nathan Eagle‟s researchThe graphic below illustrates the various data types, showing that when mobile operators see airtimeincentives, and requirements of actors in this new top-off amounts shrinking in a certain region, itdata ecosystem. tends to indicate a loss of income in that population. Such information might indicateClosing the Information Gap: Identifying increased economic distress before that datathe Returns from Better Data Use shows up in official indicators. Meanwhile, Global Pulse‟s own research into food relatedAlready, a number of organisations in the public conversations on Twitter has shown very strongand development sectors have embraced the vi correlations with food price inflation . “Thisvision of a data ecosystem in which information information comes from two brand newcaptured from these varied sources is used for sources: what people are doing and what theythe benefit of global populations. Global Pulse are saying,” says Kirkpatrick. “As a governmentis a UN initiative aimed at bringing together or aid agency, you might know that food pricesexpertise from the public, private, development, are rising or rains arent coming, but what if youand academic sectors to develop approaches could see where and how people are already Individuals Data Type: „Crowdsourced‟ information, data exhaust Sharing Incentives: Pricing/offers, improved services • Faster Outbreak Requirements: Privacy standards, „opt out‟ ability Tracking & Response • Improved Public/Development Sector Understanding of Data Type: Census data, health indicators, tax Crisis Behavior and expenditure information, facility data Data Data Mining Change Sharing Incentives: Improved service provision, increased efficiency in expenditures Commons & Analysis Requirements: Privacy standards, „opt out‟ • Accurate Mapping of ability Service Needs Private Sector • Ability to Predict Data Type: Transaction data, spending & use Demand & Supply information Changes Sharing Incentives: Improved consumer knowledge and ability to predict trends Requirements: Business models, ownership of sensitive data 4
  6. 6. changing their behaviour and prioritise where firms, governments, and individuals can beyou put resources in response?”Public health offers one of the most compellingareas where the analysis of mobile and Internet Mobilising Data to Deal with andata could lead to huge public gains. The San EpidemicFrancisco-based Global Viral ForecastingInitiative (GVFI) uses advanced data analysis In the wake of Haiti‟s devastating 2010on information mined from the Internet to earthquake, researchers at the Karolinskaidentify comprehensively the locations, sources Institute and Columbia Universityand drivers of local outbreaks before they demonstrated that mobile data patternsbecome global epidemics. GVFI‟s Chief could be used to understand the movementInnovation Officer, Lucky Gunasekara, says this of refugees and the consequent health riskstechnique can successfully predict outbreaks up posed by these movements. Researchersto a week ahead of global bodies such as the from the two organisations obtained data onWorld Health Organisation that rely on the outflow of people from Port-au-Princetraditional techniques and indicators. following the earthquake by tracking theEmploying new data collection and analysis movement of nearly two million SIM cards inmethods could be a less costly, more efficient the country. They were able to accuratelymethod of developing market intelligence for large analyse the destination of over 600,000organisations like the World Bank. The Bank already people displaced from Port-au-Prince, andspends millions of dollars each year on statistical they made this information available to viianalysis of the needs of the poor . Smarter data government and humanitarian organisationscollection and analysis could free resources for use dealing with the crisis. Later that year, ain economic development efforts . viii cholera outbreak struck the country and the same team used mobile data to track theIn a time of constrained government resources and movement of people from affected zones.reduced foreign aid, the insight produced by mining Aid organisations used this data to preparemobile data offers the possibility of preventing crises for new outbreaks. The example from Haitiand targeting services to the populations that need demonstrates how mobile data analysisthem most. Yet there are serious challenges that could revolutionise disaster and emergencyneed to be addressed before the pieces of the responses.puzzle fall into place.Obstacles on the Path to the Data convinced to share data more openly.Commons Data personalisation: When individuals haveEcosystem actors, like those described above, have multiple SIM cards, it is impossible to aggregatemuch to gain from the creation of an open data data from each SIM back to the same individual.commons. Yet the sharing of such data especially This data is most useful if it can be attached tothat tied to individuals raises legitimate concerns demographic indicators, which allow the data tothat must be addressed to achieve this cross-sector tell a story about the habits of a segment of thecollaboration. population. Improved methods of tying subscriptions to demographic information are Privacy and security: As ecosystem players needed to ensure data generated by mobile look to use mobile-generated data, they face devices is as individualised as possible. concerns about violating user trust, rights of expression, and confidentiality. Privacy and Data sharing incentives: Individuals, fearing security concerns must be addressed before security and privacy concerns, often resist sharing personal data. In addition, many private- 5
  7. 7. sector firms do not see an incentive to share data In the area of individual incentives, Jana, a Boston- they regard as proprietary. Governments often based start-up, conducts market research for global cannot force contractors to share data collected organisations in over 50 countries. The company in the execution of public contracts or make all uses SMS to survey emerging market customers in government data available for use by academia, exchange for airtime, creating a financial incentive development organisations, and companies. All for consumers to overcome their concerns about players must see material benefits and incentives sharing personal data. in data sharing that outweigh the risks. Nathan Eagle, Jana‟s founder, notes that the data Human Capital: Accurate and actionable data created is useful not only to marketing organisations mining and analysis requires considerable in private-sector companies, but has extensive technical skill, and data scientists are both in development uses as well. He cites a mobile data short supply and expensive to employ. GVFI‟s analysis effort in the huge slum of Kibera, outside Gunasekara notes that even many large Nairobi, that was used to map population change corporations do not have access to the type of and direct latrine and water pipe building efforts for expertise they need to develop novel data mining the benefit of the slums residents. techniques. “Most of these people want to start their own companies, not work for someone Government as Data Catalyst else,” he says. Maximising the contribution of Several forward-thinking governments in the human capital requires incentives for these developing world are demonstrating how individuals to use their talents for the public good government can catalyse the development of this along with long-term efforts to grow the talent ecosystem through the opening of its own datasets pool. and the active management of their dissemination and use.Overcoming the Obstacles:Novel Approaches In July 2011, Kenya launched its new Open Data Portal, which includes a full digital edition of theA number of organisations are already working to 2009 census, 12 years of detailed governmentovercome the challenges and create the incentive expenditure data, government household incomestructures needed for cross-sector cooperation. surveys, and the location of schools and healthGlobal Pulse is creating a network of Pulse Labs facilities. The portal provides unlimited data accessthat bring together experts in government, on the web and through mobile phones toacademia, the development sector, and private researchers, web and software developers,companies to pioneer new approaches to using data journalists, students, civil society and the generalfor development challenges. public. Civic organisations, mobile application developers, and media groups are already using theThe organisation is also now actively engaging with data to improve understanding of populationpartners around what Robert Kirkpatrick calls „data patterns, increase the transparency of governments,philanthropy‟, where corporations are encouraged to and map public services.share anonymised data for use by the public sectorto protect vulnerable populations. These companies The World Bank has provided support for theare driven partly by a recognition that more effective initiative, but Chris Finch at the World Bank notespolicy action will lead to greater resilience fromeconomic shocks, and therefore translate into better that the Kenyan authorities, with support at thebusiness continuity. Athletic apparel company Nike highest levels of government, drove the initiativehas demonstrated an approach to corporate data forward. In part, the Kenyan government respondedsharing through its GreenXchange patent-sharing to the country‟s growing information technologysystem. Nike is among the first corporations to sector and the new constitution‟s guarantees ofexplore opening up data publicly, and plans to share access to information. Finch notes that thedata on the sustainability of its operations. government moved forward with the Open Data Portal before the legislative, policy and legal 6
  8. 8. framework, including protections for data reuse, providing the data online, governments need towere fully in place, given momentum around invest in applications that make the data accessibleConstitutional guarantees for openness, and useful to citizens,” she says. Rotich cites the citytransparency, and participation. Policy frameworks of Chicago‟s Open Data portal as a notable example Government Catalyst • Enact appropriate legislation protecting end users without stifling innovation • Open data to the public (free or for purchase) in a way that allows for innovation without infringing on citizens privacy • Encourage the development of appropriate technological infrastructure and training of individuals capable of analyzing big data Private Sector Development • Once proper regulations are in place and public trust about the use of data has been gained, telecoms can compile or curate mobile-generated data for use by both profit-seeking enterprises and development organisations Public-Private Collaboration • Telecoms and governments must work together to find a way to track mobile information back to an individual, rather than a SIM • Government or Multi-lateral funded initiatives using data generated from mobile for development or government planning purposes (e.g., health, agriculture, education)for such protections are now being backfilled. of a government opening data in the context of aFinch sees the role of government as setting the robust multi-sector effort.legal frameworks governing data privacy andsecurity, and also in developing systems that allow Call to Actionvarious agencies and ministries to continually To realise the mutual benefits of creating anupdate the data they make available. The environment for sharing mobile-generated data, alldevelopment community can encourage this ecosystem actors must commit to active and openbehaviour by supporting progressive governments participation. Governments can take the lead insuch as Kenya‟s and linking them to the technical setting policy and legal frameworks that protectand financial resources they need. Unfortunately, in individuals and require contractors to make theirmany countries, governments are frequently seen as data public. Development organisations canpart of the challenge to establish a productive data continue supporting governments and demonstratingcommons. Kenya‟s example demonstrates that both the public good and the business value thatgovernment can take the lead. data philanthropy can deliver. And the private sector can move faster to create mechanisms for theJuliana Rotich of Ushahidi also notes that sharing of data that can benefit the public.governments must invest in applications that makethe data they are releasing useful. “Beyond Despite the challenges and risks, the opportunities 7
  9. 9. available to better serve individuals in emergingmarkets should outweigh these risks.  8
  10. 10. For more information please visit: World Economic Forum Mobile Financial Services Development Report www.weforum.orgmfs Global Viral Forecasting www.gvfi.org UN Global Pulse www.unglobalpulse.org Ushahidi SwiftRiver Platform ushahidi.com/products/swiftriver-platformThe World Economic Forum91-93 route de la CapiteCH-1223 Cologny/GenevaSwitzerlandTel.: +41 (0)22 869 1212© 2012 World Economic Forum All rights reserved.No part of this publication may be reproduced ortransmitted in any form or by any means, includingphotocopying and recording, or by any informationstorage and retrieval system. i www-01.ibm.com/software/data/bigdata/ ii www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns70 5/ns827/white_paper_c11-520862.html iii /ushahidi.com iv www.itu.int/ITU-D/ict/facts/2011/material/ICTFactsFigures2011.pdf v ess.santafe.edu/publications.html vi www.unglobalpulse.org/projects/twitter-and-perceptions-crisis- related-stress vii /www.nextbillion.net/blog/the-age-of-big-data viii jana.com/research/sample/, www.nextbillion.net/blog/2011/10/05/reaching-the-next-billion- through-mobile Haiti disaster example reference is at Karolinska Institutet (www.ki.se) 9

×