Big Data 4 Development


Sri Lanka has pretty good indicators, compared to many countries we work in. So if the UN is thinking of conventional indicators there’s not much to do. But if the intention is to bring in big data . . .
Lokanathan, S & Gunaratne, R. L.
Last week in Bangkok (23-26th March, 2015), at the invitation of the UN Development Group (UNDG) Asia-Pacific Secretariat, I had the opportunity to brief country heads and senior staff of UN agencies as well as from the Resident Coordinator’s office on how to leverage big data, for the data revolution needed to measure the progress in achieving the forthcoming Sustainable Development Goals (SDGs). The event was the Lessons Learnt Workshop for Countries Designing UN Strategic Development Frameworks (UNDAF) in 2015. 13 countries were represented: Bangladesh, Bhutan, Cambodia, DPRK, Indonesia, Iran, Lao PDR, Malaysia, Mongolia, Philippines, Papua New Guinea, Thailand, and Vietnam. The key point that I left with them was that National Statistical Organizations (NSOs) in developing economies are not yet set up to be the key champion for leveraging big data for development, let alone to certify standards. The UN’s role in my opinion was: to inform and catalyze the in-country discussions with examples from other countries.
Kreindler, G. & Miyauchi, Y.
Today I had the opportunity to speak to a mostly private sector audience in Tokyo, looking to leverage opportunities from geo-spatial information. The venue was at the G-Space x ICT International Symposium organized by Japan’s Ministry of Internal Affairs and Communications, the apex body responsible for ICT policy in Japan. I was invited to speak about LIRNEasia’s experience in leveraging mobile network big data for public purposes. In the subsequent panel discussion, I was asked how to enable international collaboration in such efforts. My answer was two part: the very realizable possibility of sharing technical know-how both in developing human capacity as well as the infrastructure required to analyze such data sets; and the potentially long path that must be walked to enable greater sharing of such data.
Colombo, the focus of our exploratory work on mobile network big data, is a tiny town by global standards: 550,000 people. But our analyses show that the surrounding area is tightly integrated contributing over 54 percent of the daytime population of the city, but contributing little or nothing to the services the commuters must be provided. A former Mayor once told me that he had thought of using the dormant power of the legislation that established the Colombo Municipal Council to establish tolls at the gates of the city. Appears this is not a problem limited to Colombo. Current debates about the efficiency of urban governance gravitate around the ‘fit’ between the size of the administrative boundary controlled by a city mayor or governor, and the actual number of people who live in the ‘wider functional metropolitan’ area.
Verizon is in the the news and under the gun for its use of supercookies to track mobile users. The company uses the tracking technology — alphanumerical customer codes known as supercookies — to segment its subscribers into clusters and tailor advertising pitches to them. Although Verizon allows subscribers some choices regarding the use of their information for marketing purposes, the company does not permit them to opt out of being tagged with the persistent tracking technology. Our discussion: Within the first cluster proposed by Solove, the most relevant problem is surveillance. In the context of big data, it is useful to distinguish between active and passive surveillance.
The Royal Statistics Society and the Overseas Development Institute had organized a well-attended public discussion on big data and the future of conventional government statistics. I was pleased that there was nothing very news said, from our perspective, because that shows that we are not lagging behind in this space. I found the comments by John Pullinger, the National Statistician of the United Kingdom, of significant interest given we are making a presentation to the senior officers of the Sri Lanka Department of Census and Statistics this coming Friday. One of his comments was that good professionals had to keep up with new techniques. If a doctor were to treat people with methods from the 1950s, they would be driven out of the profession.
On 16th January, 2015 at the invitation of the Sri Lanka Institution of Engineers LIRNEasia presented a public lecture in Colombo on the results of our ongoing big data for development research. The public lecture was organized by The Institution of Engineers Sri Lanka (IESL) and attracted over 40 people in person and an unknown number via streaming at the Wimalasurendra Auditorium in the IESL head office. LIRNEasia’s Founding Chair Rohan Samarajiva and Researcher, Danaja Maldeniya presented some of the initial findings of relevance to urban and transportation planning. They were joined in the Q&A by Sriganesh Lokanathan. The presentation slides are available HERE.
I was somewhat disappointed by the Modi government leaning toward the IBM vision of smart cities, where sensors would be ubiquitously placed across green-field new-build satellite cities across India. Our vision is lower cost and seeks to improve existing cities relying on citizens as the principal sensors. So I was pleased to our thinking echoed in a http://blogs.worldbank.org/ic4d/building-smarter-cities?
Partha Mukhopadhdyaya is an expert on cities, having studied them in multiple countries including China and India. He also happens to serve on our scientific advisory board. Mint carried the first part of an interesting discussion with Partha on cities. When we talk about the insights from big data for cities, we naturally get slotted into the data for “planning” box. But I’ve always been wary about planning and also talk about experimentation using near-real-time and low-cost feedback.
Our own work with big data focuses on cities. This guest editorial in the UN Global Pulse blog provides as excellent rationale for the focus on cities. In addition, it raises some areas for caution. Placing algorithms at the forefront (or even in the front-row seat) of decision-making may have potentially severe drawbacks. It’s indeed us who program algorithms, and we are exposed to a variety mistakes while programming.
Yuan Ze University is a leading private university in Taiwan. In keeping with its strategy of differentiation and internationalization, it has invited LIRNEasia to interact with its new Center of Excellence known as the Innovation Center for Big Data and Convergence. Even though Taiwan is not a country of focus for LIRNEasia, we have had considerable interactions with Taiwan academics over the years within the framework of CPRsouth. Professor Yu-li Liu was one of the founder members of the CPRsouth Board and helped us establish relationships in the Peoples Republic. Professor Yuntsai Chou of Yuan Ze in currently on the CPRsouth Board.
Bangladesh’s leading English language newspaper, the Daily Star, has carried a story on what that country could gain from analysis of mobile network big data. He said a lot of people are now talking about big data, but the current models of data analysis do not cover all the people, especially the poor. Mobile networks’ big data can be used to reach this segment, he added. Big data can give organisations access to more data than they have experienced before, and thus give them the opportunity to discover data correlations and patterns. Access to more accurate information can influence their business in many ways.
The specialized unit of the UN dealing with big data for development, UN Global Pulse, has constituted a data privacy advisory group. LIRNEasia’s chair and advisor to the big data for development team, Rohan Samarajiva, is a member. Among the other members are MIT’s Sandy Pentland and Umar Saif, Chair of the Punjab IT Board and Secretary IT of the Government of Punjab.
Reading the interview with Viktor Mayer-Schonberger from which the quotation below is taken, I was reminded of an exchange we had in Doha earlier this month, at ITU Telecom World. What do you have to say to those who are still unsure about ‘Big Data’? To them I say: The only way that you can be in power and not be ‘controlled’ by data is to first understand its power and then use it to your own advantage. Otherwise, you will always be at the receiving end in the balance of power between the ‘big guys’ who analyse data and the ‘every man’ who supplies data – whether consciously or inadvertently. The question was from Dr Shiv Bakhshi, now with Ericsson and whose association with me goes back to the 1990s.