There is no straightforward definition of data science, given the array of different applications. One common area is to describe data science as the ability to extract knowledge and insights from large and complex data sets.
Data science draws on the theory and practices of areas of mathematics, statistics, information science, and computer science. Data is of great value in the digital society, for both governments and enterprises, generating significant economic, social and scientific value from data.
Data science is not always accurate, however, and predictions can fall short of the mark. According to Analytics magazine there are six reasons that lead to data science shortfalls. These are: the vast volume of data (sometimes petabytes to zettabytes); data often being in motion (such as streaming data); the fact that data is found in many, often contrasting, forms; there is also uncertainty due to data inconsistency and incompleteness; data is also often in a state of change, leading to differing ways in which the data may be interpreted; and, finally, not all data is of the same value and there is relative importance to different types of complex data from different locations.
In addition, data science as a discipline is not standing still and even established data scientists need to learn new skills.
New skills
Data scientists are increasingly required to have new sets of skills, outside of statistical knowledge and information technology. Examples include an understanding of machine learning; natural language processing; geospatial analysis and advanced visualization. These are, incidentally, some of the modules that form part of the U.K. government’s data science training scheme.
Businesses
The role of the data scientists has become essential for many businesses. IT portal states that the position of data scientist has become one of the most desired and sought after positions for enterprises. Many businesses regard data science as a cog that turns business analytics towards monetizing data assets.
There are various examples in the business setting of data science in action. Price comparison websites, offering products like insurance, are one such example. PriceGrabber, PriceRunner, Junglee, Shopzilla, DealTime are some examples of data science shaping U.S. businesses.
Another area is with machine learning. Data science can help to map situations to outcomes, to predict outcomes in new situations. A business example is with predicting whether a customer who is presented a product (situation) will show interest in it or not (outcome). This can help woth reducing customer attrition; cross-selling products; optimizing products and pricing and predicting demand.
A third area is with setting up business dashboards, such as Help Desk Dashboards, for sentiment analysis of text comments; Product Dashboards for cross sell opportunities; and Operational Dashboards to provide operational metrics with data science alerts and calls to action.
Governments
In terms of government, in the Philippines data science forms a core part of government policy. According to the website Interaksyon a consequence of this is various courses springing up to train the next generation of students in the subject, this includes a newly established Master of Science qualification. In the U.K., the government is now operating its third year of a Data Science Accelerator Program, which aims to give analysts from across the public sector the opportunity to develop their data science skills.
In the U.S. it is common for state and local government to now have a chief data officer in place. According to Forbes, this to allow government agencies to begin to break down silos “to better share information and create greater transparency for citizens, businesses and other officials to find the data they need in a format they can use.”
With governments accessing so much data from their citizens it is important that data privacy is maintained and that data is used ethically. The British government has established a Data Science Ethical Framework for the public sector. This is founded on six core principles:
Start with clear user need and public benefit,
Use data and tools which have the minimum intrusion necessary,
Create robust data science models,
Be alert to public perceptions,
Be as open and accountable as possible,
Keep data secure.
Data privacy is currently a global concern, as st out in the Digital Journal article “Business needs to get smart about data protection.”
Simplicity of communication
While businesses and governments are set to make increased use of data science how information is communicated across to non-specialists is important. Data analysis will need to be presented to a range of audiences in order for them to understand the data science outcomes, in terms of businesses decisions or policy problems. Good data scientists are not simply effective at data collection and sorting, they need to be able to access, understand, and communicate the insights they get from data analysis.
Data science has a key role to play with governments and business. For it to work well, as the Harvard Business Review notes, there needs to be focus and support, and an understanding of the opportunities that are presented.
Essential Science
This article is part of Digital Journal’s regular Essential Science columns. Each week Tim Sandle explores a topical and important scientific issue. Last week we looked at hydrogen as a power source. Hydrogen can be produced at a low cost; it is relatively efficient to other fuels; and it is low polluting.
The week before we considered how information about how our internal body clocks influence health and disease can help with cancer treatments, especially the fact that the time of treatment plays a critical role.
