Table 3.

Operational definition of the classes.

Class	Operational Definition *
Analytics	Knowledge discovery by analyzing, interpreting, and communicating data
3A. Types of Analytics	Data Interpretation and Communication method
Descriptive	Exploration and discovery of information in the dataset [33]
Predictive	Prediction of upcoming events based on historical data [22]
Prescriptive	Utilization of scenarios to provide decision support [22]
3B. Types of Data	Type or nature of data used in the study
Web/social media data (WS)	Data extracted from websites, blogs, social media like Facebook, Twitter, LinkedIn [23]
Sensor data (SD)	Readings from medical devices and sensors [23]
Biometric data (BM)	“Finger prints, genetics, handwriting, retinal scans, X-ray and other medical images, blood pressure, pulse and pulse-oximetry readings, and other similar types of data” [23]
Big transection data (BT)	Healthcare bill, insurance claims and transections [23]
Human generated data (HG)	Semi-structured and unstructured documents like prescription, Electronic Medical Record (EMR), notes and emails [23]
3C. Data mining techniques	Techniques applied to extract and communicate information from the dataset
Regression	Relationship estimation between variables
Association	Finding relation between variables
Classification	Mapping to predefined class based on shared characteristics
Clustering	Identification of groups and categories in data
Anomaly detection	Detection of out-of-pattern events or incidents
Data warehousing	A large storage of data to facilitate decision-making
Sequential pattern mining	Identification of statistically significant patterns in a sequence of data
3D. Application Area	Different areas in healthcare where data mining is applied for knowledge discovery and/or decision support
Clinical decision support	Analytics applied to analyze, extract and communicate information about diseases, risk for clinical use
Healthcare administration	Application of analytics to improve quality of care, reduce the cost of care and to improve overall system dynamics
Privacy and fraud detection	Privacy: Protection of patient identity in the dataset; Fraud detection: Deceptive and unauthorized activity detection
Mental health	Analytical decision support for psychiatric patients or patient with mental disorder
Public health	Analysis of problems which affect a mass population, a region, or a country
Pharmacovigilance	Post market monitoring of Adverse Drug Reaction (ADR)
3E. Theoretical study	Discusses impact, challenges, and future of data mining and big data analytics in healthcare

* Most of the definitions listed in this table are well established in literature and well know. Therefore, we did not use any specific reference. However, for some classes, specifically for types of analytics and data, varying definitions are available in the literature. We cited the sources of those definitions.