In statistics and computer science, the term ‘big data’ generally refers to a collection of data extensive in terms of Volume, Velocity and Variety that requires more than the ability of commonly used software tools, but new specific methods and technologies for the extraction of values, analysis or knowledge.

The term big data has been used since the 1990s and the definition has changed over time. The first definition including the three Vs was given by Douglas Laney in 2001.

In a study published in 2016, De Mauro, Greco, Grimaldi proposed that the nucleus of the concept of Big Data includes the following aspects:

• ‘Volume’, ‘Velocity’ and ‘Variety’, to describe the characteristics of information;

• ‘Technology’ and ‘Analytical Methods’, to describe the requirements needed to make proper use of such information;

• ‘Value’, to describe the transformation of information into insights that may create economic value for companies and society.

In 2011 a McKinsey Global Institute report characterizes the main components and ecosystem of big data as follows:

• Techniques for analyzing data, such as A/B testing, machine learning, and natural language processing;

• Big data technologies, like business intelligence, cloud computing, and databases;

• Visualization, such as charts, graphs, and other displays of the data.

Big data can generate financial value across many sectors like healthcare, public administration, global personal and location data, retail and find, mainly with a growth of the specific economic value, growth in annual productivity and also with a reduction in working capital.

Computer and electronic products and information sectors, traded globally, stand out as sectors that have already experienced very strong productivity growth that are poised to gain substantially from the use of big data.

Finance and insurance and government can benefit very strongly from big data as long as barriers to its use can be overcome.

While all sectors will have to overcome barriers to capture value from the use of big data, barriers are structurally higher for some than for others. For example, the public sector, including education, faces higher hurdles because of a lack of data-driven mindsets and available data. Capturing value in healthcare faces challenges, given the relatively low IT investment performed so far. Sectors such as retail, manufacturing, and professional services may have relatively lower barriers to overcome for precisely the opposite reasons.

Big data in healthcare refers to the abundant health data amassed from numerous sources including electronic health records (EHRs), medical imaging, genomic sequencing, payor records, pharmaceutical research, wearables, and medical devices, to name a few.

Healthcare data analytics enable the measurement and tracking of population health, thereby enabling this switch. The use of big data analysis to deliver information that is evidence-based could, over time, increase efficiencies and help sharpen our understanding of the best practices associated with any disease, injury or illness. This also leads to remarkable advancements, even while reducing costs. With the wealth of information that healthcare data analytics provides, caregivers and administrators can now make better medical and financial decisions while still delivering an ever-increasing quality of patient care.

Various technologies are in use for protecting the security and privacy of healthcare data. The most widely used technologies are:

• Authentication: Is the act of establishing or confirming claims made by or about the subject are true and authentic. It serves a vital function within any organization: securing access to corporate networks, protecting the identities of users, and ensuring that a user is who he claims to be. Most cryptographic protocols include some form of endpoint authentication specifically to prevent man-in-the-middle (MITM) attacks.

• Encryption: Data encryption is an efficient means of preventing unauthorized access of sensitive data. Its solutions p. 3protect and maintain ownership of data throughout its lifecycle from the data center to the endpoint and into the cloud. Encryption is useful to avoid exposure to breaches such as packet sniffing and theft of storage devices.

• Data Masking: Replacing sensitive data elements with an unidentifiable value, but is not truly an encryption technique so the original value cannot be returned from the masked value. It uses a strategy of de-identifying the datasets or masking personal identifiers.

• Access Control: Once authenticated, the users can enter an information system but their access will still be governed by an access control policy which is typically based on the privilege and right of each practitioner authorized by the patient or a trusted third party.

More than ever it is crucial that healthcare organizations manage and safeguard personal information and address their risks and legal responsibilities in relation to processing personal data, to address the growing thicket of applicable data protection legislation. Different countries have different policies and laws for data privacy (such as GDPR for the EU).

As the concept of big data and data science has changed over the last few decades, the definitions of artificial intelligence (AI) have also changed over the years. Professor John McCarthy (Stanford University) in 2004 in his paper of Q&A about AI offers us the following definition in simple words: “It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.”

AI is a constantly evolving field in its applications, and the most used today are:

• Logical AI – the program decides what to do by inferring that certain actions are appropriate for achieving its goals;

• Search – AI programs examine possibilities and discoveries are made about how to do this more efficiently in various domains;

• Pattern recognition – the program compares what it sees with a specific or different pattern in several observations;

• Representation – AI uses usually mathematical logic to represent facts or data in some way;

• Inference – from some facts, others can be inferred. Mathematical logical deduction is adequate for some purposes, but also non-monotonic inferences have been added for different purposes;

• Common sense knowledge and reasoning;

• Learning from experience – the approaches to AI based on connectionism and neural nets are specialized. The limitation is that not everything can be represented correctly as useful information for programs and receivers;

• Planning – this is used for making strategies for achieving specific goals;

• Epistemology – this is a study of the kinds of knowledge that are required for solving problems in the world;

• Ontology – this is the study of the kinds of things that exist;

• Heuristics – it is a way of trying to discover an idea or a fact imbedded in a certain program;

• Genetic programming – AI solve tasks by mating random LISP (list programming) programs and selecting which fit the best.

There are already a number of research studies suggesting that AI can perform as well as or better than humans at key healthcare tasks, such as diagnosing disease. Today, algorithms are already outperforming radiologists at spotting malignant tumours, and guiding researchers in how to construct cohorts for costly clinical trials.

Most of these technologies have immediate relevance to the healthcare field, but the specific processes and tasks they support vary widely.

1. Machine learning – neural networks and deep learning. Machine learning is a statistical technique for fitting models to data and to ‘learn’ by training models with data. The most common application of machine learning is precision medicine: predicting what treatment protocols are likely to succeed on a patient based on various patient attributes and the treatment context.

2. Natural language processing – the dominant applications of NLP involve the creation, understanding and classification of clinical documentation and p. 4published research. NLP systems can analyse unstructured clinical notes on patients, prepare reports, transcribe patient interactions and conduct conversational AI.

3. Rule-based expert systems – these are based on if–then rules. Expert systems require human experts and knowledge engineers to construct a series of rules in a particular knowledge domain. However, when the number of rules is large and the rules begin to conflict with each other, they tend to break down. Moreover, if the knowledge domain changes, changing the rules can be difficult and time-consuming.

4. Physical robots – mostly for surgeries. Surgical robots, initially approved in the USA in 2000, provide ‘superpowers’ to surgeons, improving their ability to see, create precise and minimally invasive incisions, stitch wounds and so forth.

5. Robotic process automation – RPA doesn’t really involve robots but computer programs on servers. They are mostly used for administrative tasks like updating patient records or billing. When combined with other technologies they can extract data, recognize images and so on.

6. Diagnosis and treatment applications – there are also several big companies investing specifically in diagnosis and treatment recommendations for certain pathologies, with programs based on rule-based systems incorporated within national healthcare systems.

7. Patient engagement and adherence applications – providers and hospitals often use their clinical expertise to develop a plan of care that they know will improve a chronic or acute patient’s health. A growing focus in healthcare is on effectively designing the ‘choice architecture’ to nudge patient behaviour in a more anticipatory way based on real-world evidence.

8. Administrative applications – a part of the daily work schedule is spent on regulatory and administrative activities. AI can be used for a variety of applications including claims processing, clinical documentation, revenue cycle management and medical records management, spending less time on a computer and more time with patients.

There is a considerable attention to the concern that AI could lead to automation of jobs and displacement of the workforce. Of course there are factors in healthcare that could limit job losses, first of all the human touch and experience that patients, especially the elderly, value greatly.

There are also ethical implications around the use of AI in healthcare. Healthcare decisions have been made almost exclusively by humans in the past, and the use of smart machines to make or assist with decisions raises issues of accountability, transparency, permission and privacy. Many AI algorithms are virtually impossible to interpret or explain. There are also likely to be incidents in which patients receive medical information from AI systems that they would prefer to receive from an empathetic clinician.

Data scientists must learn multiple programming and database languages and master advanced statistical analysis. One factor affecting worker availability is that it takes many years to become a data scientist. Most US data scientists agree that it takes on average 4.9 years.

Important knowledge areas include statistics, computer science, which encompasses an understanding of data structures, algorithms, and database systems (e.g., Hadoop); and problem formulation (i.e., the ability to formulate problems to bring about effective solutions).

Data scientists require the strongest analytical skills as well as proficiency with specialized tools such as machine learning, Apache Hadoop, and data mining. Moreover, data scientists require generalized skills such as SQL, R, and data analysis.

Machine learning skills, in particular, are becoming mandatory for data scientists for building automated decision systems that provide future predictions. The ability to mine text is also a prerequisite for working with unstructured data, particularly in healthcare, where much of the clinical data is in note format.

While a data scientist’s role requires technical and statistical expertise, the importance of data communication should not be underestimated. For example, storytelling and communicating findings are required skills for many job positions. Also, soft skills such as problem solving and the ability to work effectively within a multidisciplinary team are needed.

A data scientist can specialize in different sectors based on personal interests: business p. 5analysis, artificial intelligence, data visualization, data mining and so on.