Chapter 8: Ethics Throughout the Data Science Cycle

anonymization

act of removing personal identifying information from datasets and other forms of data to make sensitive information usable for analysis without the risk of exposing personal information

anonymous data

data that has been stripped of personally identifiable information (or never contained such information in the first place)

autonomy

in data science, the ideal that individuals maintain control over the decisions regarding the collection and use of their data

confidentiality

safeguarding of privacy and security of data by controlling access to it

cookies

small data files from websites that are deposited on users’ hard disk to keep track of browsing and search history and to collect information about potential interests to tailor advertisements and product placement on websites

copyright

protection under the law for original creative work

cross-validation

comparison of the results of a model with different subsets of the data or with the entire dataset by repeatedly breaking the data into training and testing sets and evaluating the model's performance on different subsets of the data

data breach

the act of data being stolen by a malicious third party

data governance protocols

set of rules, policies, and procedures that enable precise control over data access while ensuring that it is safeguarded

data privacy

the assurance that individual data is collected, processed, and stored securely with respect for individuals' rights and preferences

data retention

how long personal data may be stored

data security

steps taken to keep data secure from unauthorized access or manipulation

data sharing

processes of allowing access to or transferring data from one entity (individual, organization, or system) to another

data source attribution

the practice of clearly identifying and acknowledging the sources employed in the visualizations and reporting of data

data sovereignty

laws that require data collected from a country’s citizens to be stored and processed within its borders

digital divide

gap between those who have access to digital technologies, such as the internet and computers, and those who do not

encryption

the process of converting sensitive or confidential data into a code in order to protect it from unauthorized access or interception

ethics in data science

responsible collection, analysis, use, and dissemination of data

explainable AI (XAI)

set of processes, methodologies, and techniques designed to make artificial intelligence (AI) models, particularly complex ones like deep learning models, more understandable and interpretable to humans

fairness

absence of bias in the models and algorithms used to process data

Family Educational Rights and Privacy Act (FERPA)

legislation providing protections for student educational records and defining certain rights for parents regarding their children’s records

hashing

process of transforming data into a fixed-length value or string (called a hash), typically using an algorithm called a hash function

Health Insurance Portability and Accountability Act (HIPAA)

U.S. legislation requiring the safeguarding of sensitive information related to patient health

informed consent

the process of obtaining permission from a research subject indicating that they understand the scope of collecting data

intellectual property

original artistic works, trademarks and trade secrets, patents, and other creative output

k-anonymization

principle of ensuring that each record within a dataset is indistinguishable from at least k – 1 other records with respect to a specified set of identifying attributes or features

outlier detection

identification of observations that are significantly different from the rest of the data

personally identifiable information (PII)

information that directly and unambiguously identifies an individual

pseudonymization

act of replacing sensitive information in a dataset with artificial identifiers or codes while still maintaining its usefulness for analysis

regulatory compliance officer (RCO)

a trained individual responsible for confirming that a company or organization follows the laws, regulations, and policies that rule its functions to avoid legal and financial risks

transparency

being open and honest about how data is collected, stored, and used

universal design principles

set of guidelines aimed at creating products, environments, and systems that are accessible and usable by all people regardless of age, ability, or disability