Saturday, March 5, 2011

Mapping Citizens


Mapping Citizens

IBM has developed a platform that can map India’s citizenry for all data verification purposes.
  • FEBRUARY 2011
  • BY VANTIKA DIXIT
Ram Charan lives with his family in the slums next to Munirka in New Delhi. He is unemployed. Every­day he stands in the Munirka market hoping that he can get daily wage work. His wife works as a domestic help in the nearby colonies. They have three children. The eldest child had to drop out of pri­mary government school. The high school is too far from Charan’s house and they cannot afford the bus fees.
CONNECTING DOTS IBM’s software can link people to the city map and other databases.
In the 2010 Delhi budget the govern­ment announced various schemes like urban employment guarantee scheme, allocation of money for the construction of 17 new schools in Delhi, all of which could directly better the life of Charan and his family. However, when deciding on the kind of employment schemes to offer and the locations for new schools, the civic administration realized that it does not have localized information about Delhi’s citizens. Such localized information is available in separate databases, such as the ration card data­base, the electricity meter connection database, the school records, and vari­ous other disparate data sources. The challenge is to get combined insight from them.
To make any administration respon­sive, it is necessary to have fine grained information about the demographic distribution of its citizens as such infor­mation allows the administration to plan its development schemes better. IBM Research - India has developed a technology that helps integrate differ­ent data sources, does person resolution on them, and overlays people informa­tion on the map of the city. It provides an easy to use interface that can be used by the civic authorities. “Integration of data sources is a difficult problem given that name and address records can appear in different databases in dif­ferent languages. Also, since there is no standardization, the same name and address can appear different in different databases (Ram Charan, R. K. Charan, Ram Kumar). Reconciling this vari­ability and recognizing that the same person is being talked about is a tough research problem. Also, geo coding information for Indian cities is still very coarse, and therefore overlaying names and addresses on to the city map is a challenge. We provide novel solutions to these problems and demonstrate that it is possible to provide useful infor­mation about the citizenry that can be utilized by the civic administration in developing the basic infrastructure of the city and ensuring that citizens are given a better quality of life,” says Hima P. Karanam, a scientist at IBM Research – India involved in the project.
The civil administration in India never gets a unified view of its citizens. This is because data exists in silos. The ration card database is kept separate from the passport database which is kept separate from the electricity data­base. IBM’s technology can combine information from such various data­bases by linking people across them. The system can identify the correct individual and match all his or her pro­files in various databases.
To get a single view of citizens, researchers do not propose the cre­ation of large centralized databases. Rather, they want to increase the abil­ity of users to locate relevant infor­mation spread across various sources, gain timely access for authorized use of that information, and enhance the ability to quickly “connect the dots” across various related pieces of information. This is achieved by stan­dardizing the data across sources, for example, bringing data such as name into various standard components like first name, last name, and address into city, pin code, door number, street name, area, and converting words like “street”, “st”, “gali” into a single uniform representation like street. The strength of the system is in consolidat­ing various databases and presenting a people-map overlay that can be viewed at different granularity.

No comments:

Post a Comment