How India is driving the use of open government data

Launched in 2012, India’s Open Government Data (OGD) platform has enabled millions of people to access government data to build new applications, services and more recently, to train artificial intelligence (AI) models.

Through the use of open and machine-readable formats and application programming interfaces (APIs), the platform has not only democratised access to data, but also improved transparency by enabling communities to track data points, such as those related to mortality, budget and finance, population and geography.

Lydia Clougherty Jones, senior director analyst at Gartner, said enabling public visibility and oversight of government data and insights will improve citizen trust and delivery of public services in India while driving the development of new data products.

“Open data can drive digital business outcomes by providing increased access to more data from more diverse sources, providing ‘knowable’ insights that would not otherwise be discoverable without aggregation of data from multiple government data sources.

“These insights can support more robust predictive analytics, while the underlying data can serve as training data for the ever-hungry AI models while creating new economic value by potentially matching the right data to a use case or specific business outcome,” she added.

Indian software developers have been tapping the data available through the OGD platform to build applications such as the Teeka Mobile App that lets citizens track the vaccinations of children and the pregnancy of the female members in their family.

Another app, Rainbow, helps farmers make informed decisions throughout the lifecycle of their crops by providing information such as live market prices, dam water levels and local weather updates.

But for India to unlock the value of its open data initiative, there is a need to focus on high-value datasets (HVD), according to a report by India’s National Association of Software and Service Companies (Nasscom).

Open data can drive digital business outcomes by providing increased access to more data from more diverse sources, providing ‘knowable’ insights that would not otherwise be discoverable without aggregation of data from multiple government data sources

In the report, experts called for regular maintenance of a HVD list and robust data governance to ensure high data quality, along with periodic updates communicated to users.

Indeed, Shally Gupta, graduate member and impact creator at the Institute of Electrical and Electronics Engineers (IEEE), claimed that most of the data available in India is still not rich enough or of high enough quality. “One of the obstacles India will have to overcome to get an advantage in the AI race is the lack of well-labelled, rich datasets,” she said.

Gartner’s Jones said as with other open government data initiatives, India may also struggle to define, communicate and report the tangible benefits of its open data initiatives. “To overcome this challenge, India could measure the success of its open data programme by its ‘impacts’, not the ‘volume’ of datasets downloaded.”

Dattaraj Rao, chief data scientist of Persistent Systems, an IT consultancy, singled out interoperability as another challenge. Since open data comes from multiple government organisations, he said, work is needed to make it consumable through a single interface. “Having standardised ontologies will help to solve interoperability issues.”

Nasscom’s report also drew attention to privacy-preserving technologies which it said should be made available to chief data officers on the OGD platform to manage privacy and cyber security risks.

Jones noted that the apparent lack of personal data available through India’s open data initiative should alleviate privacy concerns, adding: “India’s responsibility for the security of the datasets, once downloaded, should be diluted, if not transferred, to new users and their respective collaborations.”

But should data about individuals find its way into the OGD platform, anonymising such data can help to protect personal data – but only to some extent. “Even with data anonymisation techniques, there may be a way to identify data subjects. Hence statistical techniques like differential privacy are highly recommended,” Rao said.

Indranil Bandyopadhyay, principal analyst of financial services, insurance, data science and AI at Forrester, said a clear legal framework should be implemented to avoid uncertainties regarding copyright, data privacy and openness of India’s open data efforts.

India’s proposed Digital Data Protection Bill should address such concerns, said Devroop Dhar, co-founder of Primus Partners, a consulting firm, calling for the standardisation of data formats and a cyber security framework to ensure transparency and effective data lifecycle management.

According to Nasscom, an integrated data utilisation strategy can add $450bn to $500bn to India’s GDP by 2025. This will require “cautious utilisation of datasets and its platforms that have the maximum potential for social impact, help in policy making, create new and high-quality jobs, innovations and evangelise startups and promote research and education”.