PUBLISHED JUNE 01, 2013
Tapping the data revolution
Not just official indicators. Former UN statistics chief Paul Cheung says governments should also utilise location-based information and Big Data. By Anna Teo
PRINT |EMAIL THIS ARTICLE
- PHOTO: ARTHUR LEE
'I do hope the government will review its information policy from time to time to adjust to the new reality. At this time, it is my assessment that Singapore falls quite far behind in joining the 'data revolution'.'
- Prof Cheung on Singapore's information regime
ENTERPRISES everywhere are awash with ever-growing data of all types - proprietary, personal and public. Real time location-based information is now routinely used to track natural catastrophes, military actions, economic trends, even population shifts, and is also widely accessible through smartphones, portable navigation devices and applications such as Google Earth.
While we weren't looking, there has been a global surge in geoinformation and Big Data. It may be a data deluge but the ability to exploit information is now a key competitive edge for not only corporations but also countries. Analytics - the detection and communication of meaningful patterns and insights in data - now drives decision-making in both commercial enterprises and governments. And in this new information environment, national statistical agencies must expand their ambit beyond official statistics to include geospatial information and unstructured data, says Paul Cheung, former director of the United Nations Statistics Division (UNSD).
Geospatial information describes the location and names of features beneath, on or above the earth's surface. Examples range from topographical features on a map to location-related information that shows, for example, land use or population density.
"We are going through a geospatial information revolution. Yet there is no mechanism internationally to talk about how to manage geospatial information and its integration with other sources of information," says Prof Cheung, who was Singapore's Chief Statistician from 1991 to 2004 and returned home in January this year after nine years in the UN job. At the same time, he points out, "we are moving rapidly into an era of unstructured data, the Big Data era", characterised by volume, velocity, variety and variability. Official statistics and geoinformation, which are structured data, are easier to deal with. But more than 80 per cent of data in most organisations, it is estimated, is unstructured, in the form of text, audio, video or images.
The need for a national information authority in each country - "an organised, accredited, mandated authority within the government that deals with the whole set of information" by taking on board geoinformation and Big Data - was something Prof Cheung promoted on the global statistical front during his time at the UN. And while he has left the job and is now professor of social policy and analytics at the National University of Singapore, the push for an integrated national statistical system is something he remains zealous about.
"Some countries are doing it already; some statistical agencies are quietly repositioning themselves to be the national information authority," he tells BT.
The United States Census Bureau, the country's leading source of economic and social data, has, to begin with, a geography division that collects geographic data, maintains a large spatial database and publishes millions of maps a year. Geography provides meaning and context to statistical data, the Bureau says. Some 60 countries, including Singapore, have begun efforts at building a national spatial data infrastructure but two in particular that have gone some way in integrating geographical and statistical information are Mexico and Brazil.
"Mexico recently did a traditional economic census, but then they did a neat thing: they had all their establishments mapped out. You can see where the different shops are selling different things," Prof Cheung says.
The 2013 meeting of the UN Statistical Commission in February featured a paper by the Australian Bureau of Statistics, based on a poll of 53 national statistical offices (NSOs) worldwide, on linking geospatial information to official statistics. "The discussion was very useful and I can see that the countries are responding to this call for information integration," says Prof Cheung, who continues to attend major statistical events globally in his capacity as an adviser to the UNSD and to NSOs in several countries.
Singapore's fledgling geospatial information initiatives, driven by the Singapore Land Authority, include OneMap, an Internet portal to deliver location-based services and data to the public.
Prof Cheung is candid in his assessment. "Singapore has very good information base but I see very little integration as yet, especially over spatial frameworks," he says.
And volunteered geographic information (VGI), where individuals voluntarily create and disseminate geographic data, hardly exists here, he notes.
"We pump so much money into urban redesign and re-engineering, and so little into building information layers over the city so that we can model or plan at locality level with greater precision. There is so little community effort in pooling information together through VGI so that the redesign can benefit from public efforts from the ground up."
The survey of geospatial capabilities within NSOs was conducted under the UN Initiative in Global Geospatial Information Management (UN-GGIM), created by Prof Cheung at the UNSD.
"Decision makers require location-based information for everything, be it during disaster relief operations or to set the agenda for development work; geospatial information forms the crux of their decisions," he says.
In the commercial world, retailers such as Starbucks and 7-Eleven tap geoinformation in deciding where to locate their outlets and stores.
While the UN had recognised the importance of cartography and maps to global activities way back in 1948, advances in technology have vastly improved the availability of geospatial data and their potential integration with other types of data. A global mechanism such as the UN-GGIM would help in creating awareness among politicians and decision-makers about the powerful analytical potential of geospatial information, and in building capability in the field. But there has to be, at the same time, a global platform with common practices and standards; NSOs and the geospatial community need also to coordinate policy approaches and promote international cooperation. Several UN-GGIM conferences and high-level forums have since been held to push the agenda.
The Chengdu UN-GGIM forum, jointly organised with China's National Administration of Surveying, Mapping and Geoinformation (NASG), was originally scheduled to take place in late April. But a strong earthquake struck nearby four days before the forum - which, ironically, was on urban hazard mapping - and the event was postponed. The list of speakers included a senior official from New York City's Office of Emergency Management, who was going to share from NYC's Hurricane Sandy experience last October on delivering reliable geospatial data during emergencies.
"We want to encourage governments to have a role in geospatial information, so as to enhance the coordination, production, dissemination and use of the info," says Prof Cheung.
He feels particularly strongly about the need to assert national data over the prevalence of Google data. "Google did the world a favour by bringing geospatial information into mass use. But it also poses a threat to national authorities," he says.
"Can Google distribute any geospatial information of a country unhindered? Where is the mandate? What is the role of a national authority versus a multinational corporate?"
Google has been under fire over its collection of sensitive private data in more than 30 nations via its Street View mapping project, and recently agreed to pay a US$7 million fine in a case brought by 38 American states.
And given the prevalence of satellites in daily life today, there is also need for a convention to regulate satellite-generated data, says Prof Cheung. "Now whoever shoots the satellite up owns the data."
The rise of VGI and open-source intelligence in the Western countries also calls for a government role, he adds.
The UN-GGIM also wants to develop a robust global geodetic reference system, in support of growing demand for point-based information.
"Increasingly, we now need to have precise point positioning, PPP. In Australia, for example, it is very important because the land is huge; when you fertilise the soil, you want to drip the fertiliser into the exact precise spot each and every time. In irrigation, you want the water to drop only on the seeds. For that, you need the PPP system."
Point-based information will become rife, he suggests. "You map what's available along this route from Point A to Point B - amenities, services, shops. That's the way information is integrated, that's now routinely being done for transportation planning. You want to look at the community characteristics when planning the transport network."
And then there's the proliferation of Big Data in recent years. Sources range from commercial (credit card transactions, mobile phone usage) and administrative (medical visits, traffic monitoring) to social media and satellite images.
At a World Bank conference on Big Data last December, Prof Cheung cited, among other data sources, a quintessentially Singaporean example: electronic road pricing.
The ERP gantries capture a lot of information on traffic flows. "You can also turn the information around - you can track, for each car, how many times it passes through ERP in a single day, the distance travelled, how much is the total ERP spending on this car, relay it back to the household expenditure survey," he tells BT.
Another Singapore organisation that is a prime example of how Big Data is being mined for business insights is MHC, a managed care and third party administrator that processes one million outpatient claims here and across the region. Its database is a treasure trove of information from which one can gain insights about the behaviour of patients and doctors.
"They process one million claims a year, so they would know, from the user side, who may be abusing the system," says Prof Cheung. "In America, this kind of data analytics to check for outliers and possible fraud is very big business."
Other examples of how Big Data are being incorporated into official statistics, "as a new way of detecting, measuring social phenomena", include Estonia's use of cell phone usage to track its visitor numbers.
"They have no border controls so they use accommodation surveys. But not every tourist stays in an official hotel, so they started tracking mobile phones - any phone that's registered outside the country and being used in the country."
In a groundbreaking policy move, the US recently made government information open and easy to access - a change that the Obama Administration said "will make troves of previously inaccessible or unmanageable data easily available to entrepreneurs, researchers and others who can use those files to generate new products and services, build businesses and create jobs".
Says Prof Cheung: "The US government clearly sees both ideological and economic benefits of its open data policy. Ideologically, it fits the president's thinking of empowering the people through more information. Economically, the US government has seen how new industries have emerged to feed on the new information. Location-based apps are so big now in the US. Location analytics are being used by businesses and governments alike. This new policy will surely add momentum to the growth of industries related to data analytics in all forms."
In comparison, Singapore has a "very restrictive information regime", he says. "I do hope the government will review its information policy from time to time to adjust to the new reality. At this time, it is my assessment that Singapore falls quite far behind in joining the 'data revolution'." A new National Information Authority here could regulate on data policies and data standards, as well as serve as the platform for releasing data for public use, he says.
Born in Hong Kong, Prof Cheung took up a Singapore government scholarship in 1972 to study at the University of Singapore, and became a citizen in 1987. He was Singapore's chief population planner, in charge of getting Singaporeans to produce more babies, before he became its top numbers cruncher.
Looking back on his tenure as UNSD director, he is happy that he has strengthened the global statistical system and advanced the statistical community - indeed, NSOs and global organisations paid him glowing tributes when he stepped down last December.
At the start of his UN job in 2004, only about 70 countries, mostly developed countries, attended the UN Statistical Commission, which sets international statistical standards and methodologies. "We have now about 140 countries attending, and also many side events," he notes. He initiated, for instance, the first World Statistics Day in 2010, which had 140 countries staging events to mark the importance of statistics.
One key milestone was an update of the 1993 System of National Accounts (SNA), the global standard for GDP and economic statistics, to the 2008 version. "That wasn't easy because of different interests" among countries, he says.
The US is in the process of switching to the 2008 SNA, with China following suit soon.
Back in Singapore, he is working on developing a programme on social analytics and research at NUS. "The aim is to bring together a multi-disciplinary group of researchers to look at all the data and try to shed new light into the social phenomena," he says.
Meanwhile, as he continues his role as an adviser to governments, he will remain a leading voice in the global statistical community as official statistics "modernises" towards a new information architecture.
"Right now, when we modernise statistics, we're only thinking about a factory upgrading its processes - better raw materials, faster production, but still the same products, still a factory. Worse still, a farmer growing the same product every year. Under the new info architecture, the products may not be the same, it may be many different kinds of things. The inputs will also be very diverse. So that's what I'll continue to push, when I go around doing this kind of show and tell. The world is changing; the statistics profession must change with it."
[email protected]
Homepage
PR Newswire
