Governments today can shuffle through unstructure data across various systems to understand citizen needs and predict trends for better citizen service delivery. Here's how
Asoke K Laha | September 2, 2014
Predictions about the future with the help of statistical modeling, machine learning and data mining have been used in society and science for long. The question, however, has always been the accuracy of the predictions. And the answer lies in choosing the parameters of current and historic data so that predictions are as close to accuracy as possible. With the advent of machine learning, ability to mine enormous amount of data in a smaller time-scale, the predictions are becoming closer to the facts. Needless to say that predictive analytics is the future.
While the recently concluded FIFA World Cup 2014 ended with lots of excitement and disappointments, joy and tears, it also gives a fair indication of how far prediction software and algorithms have come. Microsoft’s Cortana, which is a digital personal assistant, not only correctly predicted the outcomes of all the 14 first-round matches but also the semi-finals and the finals. But it failed to predict the outcome of the Netherlands and Brazil match for the third position correctly. For the record, the Dutch beat the men in yellow quite convincingly.
Predictive analytics software goes through immense amounts of data to identify patterns so that forecasting can be improved and services can be delivered in a more efficient manner. It can be an incredibly powerful tool, especially for government agencies to address the needs of the public.
Less government, better governance
Government agencies maintain enormous amounts of data in various systems and databases. Leveraging this information – about its citizens, projects, operations and processes – is important for improving public services, meeting transparency norms and efficient utilisation of resources.
There are a number of scenarios in which data can be used to support a more effective decision-making process. By examining current and past data one can predict trends, which can help improve operational efficiencies and the performance of programmes. Armed with this knowledge, government agencies could run specific schemes to help individuals become more economically successful.
Predictive analytics is still a relatively new tool for government agencies. The ministry of statistics and programme implementation (MoSPI), for instance, still collects data on paper.
The ministry is responsible for compiling statistics on national accounts as well as conducting censuses. The department is also responsible for service-sector statistics to support high-level decision-making. Within the MoSPI is the national statistics office (or the NSO). The NSO is important, as it is responsible for creating and overlooking a database needed to study and examine the impact of specific issues for various population groups within India ranging from health and literacy to education and unemployment.
Data collection at the government level is still mainly a paper-driven process. Information at the district level, for example, is still collected manually. This makes the entire data collection process time-consuming, expensive and resource heavy. Despite the limitations, there have been efforts to automate the process. Some ministries – such as the housing and poverty alleviation ministry – are now using tablets for collecting data. Various states are also promoting e-governance initiatives to digitise the records of various government programs.
Trend, set, go…
Tanuj Nandan and Gopi Chand, who have written an insightful article on the application of analytics solutions in electronic governance, raise several key points on how predictive analytics can be applied in improving processes of governance. One key process is tax administration, which has presented many difficulties for India.
Tax organisations, in general, lose a lot of money through fraud, waste and abuse of unpaid taxes. For tax agencies, there is constant pressure for improving revenue collections and reducing operational costs. Just as analytics transformed credit risk management, it has the same potential for tax collection. By providing actionable information, analytics can help tax collectors reclaim more money and at the same time prevent fraud.
Another area where analytics can help is the subsidy programmes, which often fail to reach rural areas and real beneficiaries. Delivery can be improved by giving households a smart card, similar to an ‘electronic purse’, holding food stamps and education vouchers. Resources can be delivered to these smart cards directly without any bureaucratic intervention. By using analytics, decisions can be made in improving public goods and help optimise subsidy programmes.
Predictive analytics can be applied to the transportation sector as well. By analysing and predicting traffic patterns and growth, it can help improve transportation planning leading to decongestion of roads and highways.
Analytics can also be applied towards public safety and law and order. With the volume and variety of data increasing by the minute, solving crime and preventing it has become more complex. Government agencies must be equipped in handling and managing numerous data types. By using predictive analytics, security analysts are able to detect patterns to suggest likely security threats or criminal activities and, in turn, relay this to the field so that officials are constantly kept informed of current conditions.
In the United States, predictive analytic products are becoming increasingly popular with law enforcement agencies as they help officials forecast ‘hot spots’ based on the time and location of previous crimes, combined with incident reports. Some claim that with these pre-crime detection technologies, one can predict when a crime will be committed before it actually happens. Other programmes seek to analyse behavioral patterns associated with criminal or terrorist activities.
Health departments can also benefit a great deal from predictive analytics. During the H1N1 pandemic, the centre of disease control and prevention and the national institute of health, both based in the US, used predictive analytics to track the spread of the virus in order to provide information for public health advisories and activities.
By using the enormous wealth of data in their hands and converting it into actionable information, governments have the ability to look towards a future where citizen services and national security improve manifold.
Laha is president and managing director of Interra Information Technologies.
This story first appeared in Magazine Vol 05 Issue 15(01-15 Sept 2014)
The National Human Rights Commission (NHRC) has taken suo motu cognizance of media reports about the alleged inaction by the authorities of the Banaras Hindu University (BHU) into a complaint of eve-teasing and subsequent use of unwarranted manhandling and thrashing of agitating students, mostly women, by
Shipping Corporation of India Ltd (SCI) has taken delivery of a 2001 Japanese built second hand Very Large Gas Carrier (VLGC) on September 14 at Khor Fakkan in UAE. It has been renamed as Nanda Devi. The corporation’s total fleet strength stands at 70 vessels, aggregating to dea
When was the last time you saw all rival telecom corporate leaders come together. May be at time of the formal launch of prime minister Narendra Modi’s Digital India programme in July 2015. Telecom’s top brass will again share the stage at the inauguration of the three-day India Mobile Co
RailTel has achieved highest ever turnover of Rs 903 crore and clocked a profit before tax (PAT) of Rs 194 crore and net profit of Rs 130 crore in financial year 2016-2017. RailTel has said that the past year has been remarkable for it as the railway PSU has proved itself to be a key
From 1969, following an amendment in the Banaras Hindu University (BHU) Act, the university’s executive council is entirely nominated by the president of India on the recommendations of the HRD ministry. None of the executive council members are elected, which dilutes the democratic structure of the
The highest proportion of policymakers with not much or no knowledge of the Sustainable Development Goals (SDGs) was found in India (38%), said a report by an international advocacy group. Equal Measures 2030, an advocacy group, which spoke to policymakers in India, Colombia, Indonesia, Keny