 "Govt's digitisation drive may be crowd-sourced, citizens roped in")
By deploying this first-of-its-kind crowd-sourcing model, the government is trying to side-step privacy concerns that might arise out of outsourcing the task to an external data entry agency.
A government official privy to the project said the department of electronics and information technology hadn't taken a decision on the project yet, adding it was still "conceiving the project" and figuring its workability.
Also Read
If it goes ahead, this is how the plan will work: Software will be developed to convert each document into hundreds of small images. Each image will be sent to two individuals on their mobile phones. The individuals have to type the image into text (much like captchas on websites). If the entries from the two individuals match, the image on the document will be replaced by the text sent by them. As such, the document will be digitised or made machine-readable.
Since coming to power in May last year, the National Democratic Alliance government has relied heavily on crowd-sourcing ideas and inputs for its myriad projects. For instance, it has sought inputs through mygov.in, its citizen engagement platform, to decide the names of key projects such as Jan Dhan Yojana and Swachh Bharat Abhiyan, apart from suggestions for e-greetings sent by Prime Minister Narendra Modi and topics for discussion on Modi's radio talk show, Mann ki Baat, with US President Barack Obama.
The digitisation project, if and when rolled out, will perhaps put India at the fore of a group of countries such as the US that are trying to use crowd-sourcing to resolve complex problems and garner support from unusual quarters. Todd Park, a former US chief technology officer who initiated several government crowd-sourcing projects, had told Mashable in an interview, "I think (prizes and competitions) are a very exciting new tool that the government has in its toolkit to get better results at a lower cost."
The Centre's project is different from crowd-funding, a popular concept through which businesses or individuals seek monetary support for various purposes, such as funding business or education.
On incentives for citizens helping the government in this initiative, another official involved with the project said, "Students, housewives, young professionals on the go, anyone can enrol for the programme, in return for points that can later be converted into cash." As random images will be sent to individuals, it will be impossible for them to figure who a document belongs to, etc. "No one will be digitising their own records," he said, adding people would be picked on a random basis. As computers would break the documents into images and, subsequently, replace those with text, it would be at a fast pace, the official added.
"Without digitisation, the records are of no use, as all these are pictures and no meaningful analysis can be done on those," said the first official quoted earlier. On why the government needed to digitise records, as well as the privacy concerns involved, the official said such analysis could help the government better understand its own people and their records and, consequently, lead to better planning. "Why should anyone have any problem with that? We are not touching anyone's personal data such as health records. We are just digitising our own legacy data which, so far, was lying in physical registers," the official said. If records are digitised, these could also be used by other agencies for verification, before offering various services.
He explained this data could help the government in many ways. "For instance, if a new road has to be laid, the local agency responsible could quickly get information on the farmlands that will be impacted, their owners and their record of rights, and plan accordingly."
Outsourcing the entire project could have exposed the government and citizens to the risk data theft, said the second official. Also, the government would have to put in place considerable resources to cross-check digitised records. "As the image will be sent to two random individuals, the chances of errors become very small. Even then, if we feel there are mistakes creeping in, we could make it stronger by sending it to three people," said the official.
Through the past few years, various arms of the government have tried to digitise their records, though most of these are lying in silos and haven't been centralised. If the government goes ahead with the Digital Locker initiative, the exercise will involve five to six billion records (assuming each citizen uploads four to five documents). The beta version of Digital Locker is already being used by 105,852 people and the final version is expected to be launched in the next couple of months.
Neel Ratan, executive director of audit and consultancy firm PwC, said there were multiple ways in which authentic digital records could be availed of. "The first is to get these from the issuer department, as most agencies have already digitised their records in some way. This will be the most authentic manner," he said. Another way was roping in a trusted third party to verify the scanned image with the originals, he said. "The crowd-sourcing method could have limited benefits, as you only get digitised copies for usage, but it will not verify the authenticity of the scanned documents with the originals. Other methods could be more efficient in determining the authenticity," added Ratan, who is working closely with the government on the Digital Locker project.
- Users upload scanned copies of their records on the digital locker
- The scanned record is broken down into small images by a software
- One image sent to two random individuals on their mobile phones
- Individuals convert image into text like a captcha and send it back
- Software matches the two entries and replaces the image with text on the document
- This procedure is repeated for all documents